+ All Categories
Home > Documents > Motivation and Its Relationship With Reading Achievement in an Urban Middle School

Motivation and Its Relationship With Reading Achievement in an Urban Middle School

Date post: 04-Feb-2017
Category:
Upload: jonah
View: 228 times
Download: 3 times
Share this document with a friend
22
This article was downloaded by: [130.108.121.217] On: 26 August 2014, At: 11:01 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK The Journal of Educational Research Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vjer20 Motivation and Its Relationship With Reading Achievement in an Urban Middle School Norman Unrau a & Jonah Schlackman a a Charter College of Education California State University, Los Angeles Published online: 07 Aug 2010. To cite this article: Norman Unrau & Jonah Schlackman (2006) Motivation and Its Relationship With Reading Achievement in an Urban Middle School, The Journal of Educational Research, 100:2, 81-101, DOI: 10.3200/JOER.100.2.81-101 To link to this article: http://dx.doi.org/10.3200/JOER.100.2.81-101 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Transcript
Page 1: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

This article was downloaded by: [130.108.121.217]On: 26 August 2014, At: 11:01Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

The Journal of Educational ResearchPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/vjer20

Motivation and Its Relationship With ReadingAchievement in an Urban Middle SchoolNorman Unrau a & Jonah Schlackman aa Charter College of Education California State University, Los AngelesPublished online: 07 Aug 2010.

To cite this article: Norman Unrau & Jonah Schlackman (2006) Motivation and Its Relationship With Reading Achievement inan Urban Middle School, The Journal of Educational Research, 100:2, 81-101, DOI: 10.3200/JOER.100.2.81-101

To link to this article: http://dx.doi.org/10.3200/JOER.100.2.81-101

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

hrough the perspective of self-determination the-ory (Deci & Ryan, 1985, 2002; Ryan & Deci,2000), the authors explore motivation to read in

an urban middle school. Self-determination theory (SDT)provides a research-based perspective of motivation, espe-cially motivation related to school, and to literacy growthin particular. SDT emphasizes the central importance ofeach individual’s need for personal development and self-regulation. Although researchers have conducted extensiveinvestigations of intrinsic and extrinsic motivation inschool contexts (Deci & Ryan; Harter, 1981), fewresearchers have examined the trajectory of intrinsic andintrinsic motivation to read, especially across grades, gen-der, and ethnicity (Lepper, Sethi, Dialdin, & Drake, 1997;Wang & Guthrie, 2004).

Intrinsic Versus Extrinsic Motivation

Educators who developed SDT (Deci & Ryan, 1985;Ryan & Deci, 2000) made distinctions between intrinsicand extrinsic motivation on the basis of extensive inves-

tigations. Intrinsic motivation arises from an individual’spersonal interest in a topic or activity and is satisfiedthrough pursuit of that topic or activity. Intrinsic motiva-tion, central to self-regulated learning and self-determi-nation, embodies a student’s desire for mastery, sponta-neous curiosity, and inquiry. Often an intrinsicallymotivated person enters the flow of experience related toan interest or activity (Csikszentmihalyi, 1990). Forexample, an intrinsically motivated middle school readerwho was enthralled by the Harry Potter series (Rowling,1998) would find internal satisfaction in reading a bookin the series—entering Harry’s world and participating inhis magical adventures. Conversely, extrinsic motivationarises from participation in an activity, not for its ownsake, but for rewards or the release from some externalsocial demand. An extrinsically motivated middle schoolreader who faces a quiz on a chapter in an assigned read-ing and wants to receive a high grade while avoiding arebuke from a teacher or parent would read to performwell on the test.

Deci, Koestner, and Ryan (2001) have argued fromextensive meta-analyses of relevant studies that rewardsassociated with extrinsic motivation can undermine intrin-sic motivation. However, consistent with SDT, especiallyits emphasis on intrinsic and extrinsic dimensions of moti-vation (Lepper & Henderlong, 2000), student academicreading is probably driven by a blending of intrinsic andextrinsic motivation.

Although extrinsic motives or reward systems, such asgrades, may heighten students’ perceptions that the locusof control for their actions is external, not all forms ofextrinsic motivation lead inevitably to the underminingof intrinsic motivation. Some forms of extrinsic motiva-tion may become internalized and “owned” by the stu-dent. According to SDT, extrinsic motivation can take

Address correspondence to Norman Unrau, Division of Curriculumand Instruction, Charter College of Education, California State Uni-versity, Los Angeles, 5151 State University Drive, Los Angeles, CA90266. (E-mail: [email protected])

Copyright © 2006 Heldref Publications

81

Motivation and Its Relationship With Reading Achievement in an Urban Middle School

NORMAN UNRAUJONAH SCHLACKMANCharter College of EducationCalifornia State University, Los Angeles

ABSTRACT The authors investigated the effects ofintrinsic and extrinsic motivation on reading achievementfor urban middle school students. The authors initiallydeveloped a research-based theoretical model representinginterrelationships among students’ ethnicity, gender, gradelevel, intrinsic motivation, extrinsic motivation, and read-ing achievement. When the authors tested the model, struc-tural equation models (SEMs) produced effects that werestatistically significant. For example, a multiple-groupsSEM analysis indicated that intrinsic motivation had astronger positive relationship with reading achievement forAsian students than for Hispanic students. Analysis by gen-der, ethnicity, and grade level showed statistically signifi-cant decreases over time for intrinsic and extrinsic motiva-tion. Results suggest a significant decline in overallmotivation to read during the middle school years for stu-dents in urban schools.

Key words: cross-cultural differences, intrinsic and extrinsicmotivation, reading ability or achievement

TDow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 3: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

four different forms, ranging from motivation in personswho are less self-determined to motivation in those whoare potentially high in self-determination. The perceivedloci of causality include (a) external regulation (leastautonomous), (b) introjected regulation (motive andassociated behavior not fully embraced as one’s own), (c)identified regulation (valuing of a goal or regulation andassociated behavior as personally important), and (d)integrated regulation (most autonomous extrinsic moti-vation that is fully assimilated into the self and one’s ownvalues but not inherently enjoyed). The last two forms ofextrinsic motivation, identified and integrated regula-tion, are most likely to become adopted by young readersas inherently meaningful or valuable. Many studentsinternalize getting good grades as a personally meaningfuland valuable pursuit.

In general, researchers have repeatedly found that stu-dents’ motivation declines as they progress from elemen-tary school through high school (Eccles, Lord, &Buchanan, 1996). In the early 1980s, Harter (1981) foundthat students in Grades 3–9 reported gradual declines inintrinsic motivation. Gottfried, Fleming, and Gottfried(2001) also documented declining intrinsic motivationfrom Grades 4 through the high school years. However,they found that declines in some content areas were moreprecipitous than others: (a) mathematics and science hadthe largest declines, (b) reading had more modest deterio-ration, and (c) social studies revealed no significant drop.

In a recent study of intrinsic and extrinsic motivationamong students in Grades 3–8, Lepper, Corpus, and Iyen-gar (2005) found that intrinsic motivation deterioratedsignificantly. Incidentally, students’ grades and standard-ized test scores correlated positively with measures ofintrinsic motivation. However, extrinsic motivationchanged little through the grades and correlated nega-tively with academic performance: the higher students’extrinsic motivation, the lower their academic perfor-mance. In a 3-year longitudinal study of students pro-gressing from middle to high school, Otis, Grouzet, andPelletier (2005) reported that students’ intrinsic andextrinsic motivation decreased gradually as studentsmoved from Grade 8 to Grades 9 and 10.

Many educators have speculated about the reasons forthe weakening of school-related motivation, includingthe notion that efforts to motivate students through var-ious reward systems undermine their own interests andnatural curiosity (Kohn, 1993). Although the reasons fora general decline in motivation are complex, we furtherexplore our interests in motivation for reading in partic-ular. More information can be learned about the relation-ship between intrinsic motivation, extrinsic motivation,reading achievement, and other variables, such as gradein school, gender, and ethnicity. The importance of gainsin knowledge in these areas is particularly critical in themiddle school years because these years serve as a step-pingstone to high school and beyond.

McKenna, Kear, and Ellsworth’s (1995) Model of Reading Attitude

Drawing on the theoretical models and research ofMatthewson (1976), Liska (1984), and others, McKenna,Kear, and Ellsworth (1995) postulated their own model ofchanges in reading attitude, in which attitude is viewed asaffective and beliefs are viewed as causally related to atti-tude. In accord with the model developed by McKenna andhis colleagues, the social structure and environment ofreaders directly affect their beliefs and intention to read.Social context and environment shape readers’ beliefsabout others’ expectations in response to readers’ motiva-tion to conform to those expectation and their beliefsabout the outcomes of reading in light of their opinionsabout the value of these outcomes. The beliefs aboutexpectations and outcomes, in turn, influence readers’ sub-jective norms, especially those shaped by the expectationsof significant others and attitude toward reading. The sub-jective norms are contingent on factors such as (a) cultur-al identity of others, (b) degree of identification with thatculture, (c) specific expectations, (d) imposed purposes forreading, and (e) conflicts among expectations. Meanwhile,attitude toward reading is often contingent on factors suchas personal purpose for reading, intensity of purpose, andimportance of the reading. Intention to read often dependson readers’ interest in the topic, physical environment, andavailability of support. The decision to read or continuereading is subsequently determined not only by subjectivenorms and attitude toward reading but also by the inten-tion to read.

McKenna and colleagues (1995) suggested that predic-tions about reading attitudes can be hypothesized withtheir theoretical model, which relies on earlier theoreticalmodels, as well as on previous experimental research.According to their theory, if a reader is frequently frustrat-ed when reading, those frustrating experiences can con-tribute to a belief that reading is a frustrating experience.Consequently, that reader’s attitude toward reading wouldbecome more negative. Conversely, a reader who findsenjoyment in reading experiences would develop a beliefthat reading is enjoyable and that reader’s attitude towardreading would become more positive. For readers whosebeliefs and attitudes toward reading become more negative,alternative modes of entertainment, such as video games,would vie for the student’s reading time and contribute tothe gradual deterioration of not only reading time but alsoreading skills and the growth of knowledge.

McKenna and colleagues (1995) investigated develop-mental trends in recreational and academic reading atti-tude in the elementary grades and in the relationshipbetween those attitudes and the variables of reading abili-ty, gender, and ethnicity. The researchers found that recre-ational and academic attitudes began at a relatively posi-tive level in the first grade and ended in “relativeindifference” (p. 952) in the sixth grade. They also found

82 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 4: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

that the increasingly negative attitude toward recreationalreading was related to ability, with least able readers’ atti-tudes declining most rapidly. That is, as children progressedthrough the elementary grades, those with less abilityrevealed a more rapidly declining interest in reading.Unfortunately, an increasingly negative attitude towardacademic reading grew regardless of readers’ ability. Theresearchers believed that the negative trend in reading atti-tude could be attributed to the impact of weakening read-ing ability in the upper grades. Girls had more positive atti-tudes than did boys at every grade level for both forms ofreading; however, the differences between boys and girlswere unrelated to ability. McKenna and colleagues foundthat in their study involving African American, White,and Hispanic students, ethnicity contributed little to thenegative attitude trend in either form of reading. Theresearchers believed that the absence of meaningful effectsof ethnicity on reading suggest that cultural norms for read-ing are similar in the three cultures included in that study.They also suggested that the lack of an effect for ethnicitywas rooted in their sample, which may not have includedsufficient representation from inner-city schools with largenumbers of minority children.

Findings by McKenna and colleagues (1995) suggest thatresearchers should further investigate trends in attitudetoward reading and their likely affect on reading abilitythrough the middle school years. Such research couldexpand knowledge of attitudes toward reading as childrenprogress through middle school, especially attitudes amongurban middle school students whose weakening motivationfor literacy and performance may lay the groundwork fordropping out during the early high school years. Perhapsmore important, McKenna and colleagues’ finding thatethnicity has no meaningful effects on attitude towardreading should be compared with findings from urbanschools with more students from diverse cultures. Ifresearchers would implement that research by studyingurban middle schools with large numbers of minority stu-dents from different ethnic groups, educators could discov-er whether there are meaningful differences across ethnici-ties in reading attitude and performance. Educators couldalso discover what becomes of gender differences in readingmotivation among minority children in urban areas as theymove through middle school.

Development of the Motivation for Reading Questionnaire(MRQ)

Work on literacy engagement in the elementary schools,including that done by McKenna and colleagues (1995)and by the National Reading Research Center (NRRC;Guthrie, McGough, & Wigfield, 1994; Wigfield &Guthrie, 1995, 1997; Wigfield, Guthrie, & McGough,1996) during the 1990s, serves as a base for further discov-eries about middle school reading engagement. Researchon motivation at NRRC was grounded in an engagement

perspective integrating cognitive, motivational, and socialaspects of reading with achievement motivation theory,which includes readers’ competence and their beliefs abouttheir own efficacy as readers, as well as intrinsic and extrin-sic motivation and purposes for achievement. Using thosedomains of research and theory as resources and buildingon previous related investigations, Guthrie, Wigfield, andassociates (Guthrie et al., 1994; Wigfield & Guthrie, 1995,1997; Wigfield et al., 1996) developed the Motivation forReading Questionnaire (MRQ), which consists of 11dimensions of reading motivation. Using fourth- and fifth-grade students, Wigfield and Guthrie (1997) refined theMRQ to 55 items that measured 11 dimensions: (a) self-efficacy, (b) challenge, (c) work avoidance, (d) curiosity,(e) involvement, (f) importance, (g) recognition, (h)grades, (i) competition, (j) social motives and (k) compli-ance. In the fall and spring semesters, Wigfield and Guthrie(1997) obtained reliability measures that indicated thedegree to which items in each scale were related. Whereassome measurements of the internal consistency reliabilitiesfor a scale were at or above an acceptable .70, others wereless consistent, indicating that items within those scaleswere less closely related.

Combining theoretical and empirical perspectives, Wig-field and Guthrie (1997) created intrinsic and extrinsiccomposites of reading motivation based on the MRQscales. Initially, the intrinsic composite included the effica-cy, curiosity, and involvement scales, and the extrinsiccomposite was made up of the recognition, grades, andcompetition scales. The researchers subsequently foundthat their intrinsic composite predicted the amount andbreadth of reading better than did their extrinsic compos-ite. In a follow-up investigation of the contributions thatmotivational variables make to amount of reading, readingachievement, and text comprehension, Guthrie, Wigfield,Metsala, and Cox (1999) formed an intrinsic compositeconsisting of the challenge, curiosity, and involvementscales. They also formed an extrinsic composite that con-sisted of the recognition and competition scales. Theyfound that their intrinsic and extrinsic composites con-tributed substantially to predicting amount of readingwhich, in turn, facilitated text comprehension.

Building on that base of prior research, Baker and Wig-field (1999) found that with a substantially larger sample offifth- and sixth-grade students, the 11 dimensions of read-ing in the MRQ reliably measured motivation for reading.Using confirmatory factor analysis (CFA), Baker and Wig-field demonstrated that the proposed 11 dimensions ofchildren’s reading motivation could be identified and mea-sured reliably. Baker and Wigfield (1999) also exploredgrade, gender, ethnicity, amount of reading, and readingachievement. In an earlier study, Wigfield and Guthrie(1997) found significant motivational differences betweenstudents in Grades 4 and 5 that favored fourth-grade stu-dents. In their subsequent study, Baker and Wigfield (1999)found that fifth-grade students had significantly higher

November/December 2006 [Vol. 100 (No. 2)] 83

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 5: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

scores on the social and recognition scales of the MRQthan did sixth-grade students. Results of the separate stud-ies indicate deterioration of motivation to read as childrenprogressed through the late elementary grades. Statisticallysignificant gender differences also appeared in the stud-ies—girls scored higher on all scales except the competi-tion and work-avoidance scales.

Because of limited research and gaps in knowledge aboutreading motivation across ethnic groups, Baker and Wig-field (1999) also investigated ethnicity differences in read-ing motivation. Their sample of about 300 students con-sisted of approximately 50% White and 45% AfricanAmerican students. African American children had highermean scores than did White children on all scales exceptcuriosity, social, and work avoidance. All of the scales werepositively related to children’s reports of reading activityand to reading achievement, whereas the relation betweenreading motivation and achievement was greater for girlsand for White students than for boys and African Ameri-can students.

The investigation that Baker and Wigfield (1999) com-pleted demonstrated that reading motivation is multidimen-sional. However, the question of what and how many dimen-sions constitute reading motivation for which students inwhat grades remains open to debate. Furthermore, Baker andWigfield called for future research using samples from differ-ent populations. Examining relationships between theMRQ, especially its intrinsic and extrinsic composites, andvariables such as grade in school, gender, and ethnicitywould clarify one’s understanding of motivational patterns aschildren move through the middle school years.

Examining Ethnicity Differences in Reading Motivation

Researchers have gathered evidence (Chen, Stevenson,Hayward, & Burgess, 1995; Graham, 1994; Stevenson,Chen, & Uttal, 1990; Thernstrom & Thernstrom, 2003)and argued (Obgu, 1981; Steinberg, 1996; Thernstrom &Thernstrom) that cultural values and beliefs can shape stu-dents’ motivation to undertake academic tasks. Accordingto Thernstrom and Thernstrom, culture matters andexplains differences in the academic performance of Asian,Hispanic, African American, and Caucasian students.Although few studies explore the effects of culture on stu-dents’ motivation in reading, investigations of those rela-tionships hold the promise of gaining deeper understand-ings of the important interactions between ethnicaffiliations and reading behavior. McKenna and colleagues’(1995) model acknowledges the potential of a culture toinfluence attitude toward reading through subjectivenorms, but they did not find statistically significant differ-ences attributable to ethnicity. Our study helps to close theliterature gap regarding relationships between culture andreading behavior. Specifically, we investigate (a) differ-ences in intrinsic and extrinsic motivation among ethnic

groups and (b) relationships between motivation and read-ing achievement among ethnic groups.

Building on ways in which ethnicity and culture mightaffect motivation, Wang and Guthrie (2004) extendedmotivational research using the MRQ to discover thedegree to which intrinsic motivation and extrinsic motiva-tion correlated with text comprehension between U.S. andChinese students. They wanted to determine whether stu-dents’ motivation for reading comprehension tasks was dif-ferent between the two cultures. Based on (a) research byWigfield and Guthrie (1997) and Baker and Wigfield(1999), (b) subsequent exploration of motivational com-posites (Guthrie et al., 1999), and (c) their own explo-ration of these composites, Wang and Guthrie hypothe-sized and confirmed a two-factor (intrinsic and extrinsic)motivational model to predict text comprehension. Theirsample consisted of a U.S. group of fourth graders and aChinese group of fourth graders who lived in Taiwan andwere native Chinese speakers from middle-class families.(For the Chinese students, the MRQ was translated intoChinese.)

The intrinsic composite proposed by Wang & Guthrie(2004) consisted of the curiosity, involvement, and chal-lenge scales, whereas their extrinsic composite was basedon the recognition, grades, social, competition, and com-pliance scales. Wang and Guthrie found that intrinsicmotivation positively related to and predicted text com-prehension equally for U.S. and Chinese students. Extrin-sic motivation, however, was negatively related to textcomprehension for both groups.

Because the participants in our study were predominant-ly Hispanic and Asian students, we explored the relation-ships among intrinsic motivation, extrinsic motivation,grade level, gender, ethnicity, and reading achievement forthese two groups. To represent the relationships amongthose variables, we proposed a theoretical model to describedirect and indirect relationships among them (see Figure 1).Similar to research by Wang and Guthrie (2004), wehypothesized that an intrinsic motivational constructunderlies the curiosity, involvement, and challenge scalesand that an extrinsic motivational construct underlies therecognition, grades, social, competition, and compliancescales. In addition, our model further hypothesizes that stu-dents’ grade, gender, and ethnicity are likely to affect notonly intrinsic and extrinsic motivation but also readingachievement. We tested the adequacy of the theoreticalmodel with structural equation modeling (SEM), which is astatistical approach that simultaneously estimates direct andindirect relationships among the variables in the model.

Given the variations in previous theoretical models,apparent gaps in the research, and the need to clarify andextend the understanding of motivation to read, we exam-ined the following research questions regarding the relation-ships among reading motivation, reading achievement, gen-der, grade in school, and ethnicity in an urban middle school:

84 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 6: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

1. To what extent does intrinsic and extrinsic motivationrelate to the reading achievement of students in middleschool?

2. What are the relationships among gender, grade, intrin-sic and extrinsic motivation to read, and readingachievement?

3. To what extent do the variables and their relationshipsin the structural models arising from Questions 1 and 2differ across Hispanic and Asian middle school students?

4. Does intrinsic and extrinsic motivation change signifi-cantly over time for middle school students across eth-nicities, school grade, and gender?

Method

Participants

While serving as a university literacy coach to an urbanmiddle school of approximately 2,000 students in the LosAngeles area, the first author worked with faculty andadministration to understand more deeply students’ litera-cy and learning processes, including how these changedfrom grade to grade. As a member of the school’s literacy

committee formed to investigate and address literacy issuesin the school, the first author suggested administering theMRQ and the Gates-MacGinitie test to students in Grades6, 7, and 8. The purpose for administering the testing pro-gram was to develop a deeper knowledge of students’ moti-vation and its relationship to their reading performance.

Participants in the study were students in Grades 6, 7,and 8 who attended the urban middle school. The school’spopulation included about 75% Hispanic and 20% Asianstudents. The remaining students were African American,American Indian, and White students. About 75% of theHispanic students were first- or second-generation Mexi-can American; the remaining students were from CentralAmerican countries, especially El Salvador. More than80% of the Asian students lived in Chinatown, which waswithin the school’s service boundary, and were first- or sec-ond-generation Chinese. The community in which theschool was located included many businesses displayingsigns in Spanish, such as grocery stores and auto repairshops, which served the mostly Hispanic population. Themajority of the participants were from economically disad-vantaged families; over 90% of students participated in thefree or reduced-price lunches program.

November/December 2006 [Vol. 100 (No. 2)] 85

Curiosity

FIGURE 1. Theoretical model of motivation and reading achievement.

Involvement

Challenge

Grade

Gender

Ethnicity

IntrinsicMotivation

ExtrinsicMotivation

ReadingAchievement

Recognition

Grades

Social

Competition

Compliance

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 7: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

About 3% of the school’s students were English languagelearners (ELLs). However, students in English-as-a-second-language (ESL) classes were not included in this study. His-panic and Asian students were behind national averages inreading achievement. Approximately 70% of the teacherswere fully credentialed, and the remaining teachers werepursuing their state credentials through intern or other pro-grams. The average class size in the school was approxi-mately 28 students. Because of the relatively small numbersof African American, American Indian, and White stu-dents in the sample, we did not analyze these groups in sub-sequent statistical analyses. Hence, we analyzed only datafrom the Hispanic and Asian students.

Measures

Motivation for Reading Questionnaire. Wigfield and col-leagues (1996) designed the 54-item questionnaire to assess11 aspects of reading motivation. Because students in middleschool do not have reading classes, and five items referred toreading classes, we asked one of the designers of the MRQabout modifying items referring to this class. He suggestedsubstituting “English class” for reading class (J. Guthrie, per-sonal communication). Concerns about validity or reliabili-ty did not arise at the time of the discussion. Where originalitems referred to reading classes, the modified items refer toEnglish classes. Responses to the items on the MRQ rangedfrom 1 to 4 (1 = very different from me, 2 = a little different fromme, 3 = a little like me, and 4 = a lot like me).

Baker and Wigfield (1999) reported the 54 items of theMRQ and the motivational dimensions that each mea-sures. Although the MRQ has 54 items, Baker and Wigfieldeliminated 4 items from two scales—two items from thecompetition scale (Numbers 18 and 22) and two items fromthe compliance scale (Numbers 4 and 6) on the basis oftheir results of CFA. To permit fair comparisons betweenthe Baker and Wigfield study and our study, we alsodropped the 4 items (Items 18, 22, 4, and 6) from our com-petition and compliance scales. Whereas Baker and Wig-field used the MRQ with Grade 5 and 6 elementary schoolstudents, our interest was, in part, to discover possiblechanges in motivation among older children during themiddle school years.

Although researchers have used the MRQ in manystudies of reading motivation, Watkins and Coffey (2004)raised questions about its structural or factorial validity.According to them, the theoretical 11-factor structure setforth by Baker and Wigfield (1999) did not adequately fitCFA data from their two samples of elementary schoolstudents. Subsequent exploratory factor analyses yielded8 factors for each of their two samples. The confirmatoryfit index (CFI) reported by Watkins and Coffey rangedfrom .85 to .88 for their various models. The CFI, whichis a measure of model fit in CFA, ranges from 0 to 1,where a CFI = 1.00 indicates a perfect model fit in whichthe estimated model parameters perfectly explain the

observed variances and covariances among the data. Inpractice, CFIs of .90 and higher indicate a relatively goodmodel fit.

To address the issue of dimensionality raised by Watkinsand Coffey (2004), we conducted a CFA based on (a) 11 fac-tors of reading motivation specified by Baker and Wigfield(1999) and (b) 8 factors of reading motivation based on theresults of an 8-factor exploratory factor analysis (EFA).Because the main focus of our study was on students’ Time 2MRQ (and Gates-McGinitie) scores, we tested the dimen-sionality with scores from only the second administration ofthe MRQ. For those scores, an 11-factor model resulted in aslightly higher model fit (CFI = .90) than an 8-factor model(CFI = .88). We proceeded with the 11-factor structure forthe MRQ because our 11-factor model followed the samepattern of loadings proposed by Baker and Wigfield, and wewanted to compare our findings with their results. However,the CFI resulting from our 8-factor model (CFI = .88) wassimilar to the CFIs for 8 factors found by Watkins and Cof-fey. Because there is no empirically grounded consensusamong researchers regarding the number of factors underly-ing the MRQ, we suggest that researchers investigate itsstructural validity and dimensionality.

Gates-MacGinitie Reading Test. We used the Level 5/6 ver-sion of the Gates-MacGinitie Reading Test (3rd ed.;MacGinitie & MacGinitie, 1989) as a measure of readingachievement. Students in Grades 6, 7, and 8 completedeither Form K or L in the fall of the second year of the study.The alternate form reliability coefficient for Forms K and Lis reported as .90 by the creators of the test. The Vocabularysubtest contains 45 multiple-choice items of children’s read-ing vocabulary. The Comprehension subtest contains 14narrative and expository passages and 48 multiple-choicequestions based on the passages. The test is widely used andhas been reported to have good stability and internal-con-sistency reliability (Lipson & Wixson, 1997).

Procedure

We collected data for this study across 2 academic years.The MRQ was administered in the fall of Year 1 and againin the fall of Year 2, which established two grade cohorts.We labeled students in Grade 6 in Year 1 and in Grade 7in Year 2 as the “6–7” cohort; students in Grade 7 in Year1 and in Grade 8 in Year 2 were the “7–8” cohort. Theassistant principal monitored the distribution and collec-tion of the MRQ questionnaires, which were completedduring the students’ English classes. Also administeredduring the fall of Year 2 was the Gates-MacGinitie test ofreading achievement. All students in Grades 6, 7, and 8took the Gates-MacGinitie test except (a) those in Levels1–4 ESL classes and (b) those in Special Education class-es who were not mainstreamed into regular English class-es. Students took either Form K or L of the Gates-MacGinitie Reading Test (3rd ed.). Personnel in the TitleI office of the middle school scored the responses to the

86 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 8: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

Gates-MacGinitie test with materials provided by thepublisher of the test.

MRQ scale scores. The first step in the analysis of theMRQ items involved averaging subsets of the MRQ ques-tionnaire items, which we did for each subset of items rep-resenting the 11 constructs that Baker and Wigfield (1999)suggested comprise the MRQ. Hence, the MRQ items werereduced to 11 dimensions measuring (a) efficacy, (b) chal-lenge, (c) work avoidance, (d) curiosity, (e) involvement,(f) importance, (g) recognition, (h) grades, (i) competi-tion, (j) social, and (k) compliance. We reported themeans, standard deviations, and measures of internal con-sistency reliability for each of the MRQ scale scores. Cor-relations among the scale scores are also presented. Weshow the descriptive statistics in this study (e.g., means,standard deviations, and correlations) (a) for the entiresample, (b) aggregated by gender, and (c) aggregated byethnicity (Hispanic and Asian). Furthermore, we com-bined some of the MRQ scale scores to represent students’intrinsic and extrinsic motivation.

Similar to Wang and Guthrie’s (2004) study, we aver-aged the MRQ scales of curiosity, involvement, and chal-lenge to “operationalize” students’ intrinsic motivation.To measure students’ extrinsic motivation, we averagedthe MRQ scales of recognition, grades, social, compli-ance, and competition. The items comprising the MRQscales used by Wang and Guthrie differed from the workof Baker and Wigfield (1999); Wang and Guthrie includ-ed extra items on some of their scales. However, we fol-lowed the same procedures and items used by Baker andWigfield to construct each scale; we did not use the extraitems that Wang and Guthrie included in their curiosityand involvement scales. We believe that this difference isminor and should not greatly affect the measurement ofthe latent construct that the set of items is expected tomeasure.

Analysis

We conducted the analysis with two subsets of the origi-nal data. The first set of SEM analyses examined the rela-tionship between intrinsic and extrinsic scale scores andthe Gates-MacGinitie test. We included only students whohad complete data on the MRQ scales, the Gates-MacGinitie test, and information on gender, ethnicity(Hispanic or Asian), and grade in Year 2 of the study. Thatresulted in a sample size of 1,032 students (768 Hispanicstudents and 264 Asian students). The variable grade inthe SEM analyses represents the students’ grades in Year 2.Therefore, the effect of grade in the analyses should beinterpreted as a between-grade effect.

In the second set of analyses, we used a sample thatincluded students who had complete data on MRQ scoresat Time 1 and Time 2, as well as information on gender,ethnicity (Hispanic or Asian), and grade cohort. Thatresulted in a sample size of 470 students. We used differ-ences in mean scores and dependent samples t tests to testthe statistical significance of the change in motivation overtime. Those analyses helped us investigate the possibledeterioration of motivation over time and whether therewere differences in such deterioration across gender, eth-nicity (Hispanic or Asian), and grade cohort (6–7 cohortor 7–8 cohort).

Results

Table 1 shows correlations among the MRQ scales, theGates-MacGinitie test, and gender. The correlation ofgender with any of the MRQ scales and the Gates-MacGinitie test is a biserial correlation. All other correla-tions are Pearson product-moment correlations. For His-panic students, reading achievement correlated negativelywith curiosity (r = –.11, p < .01), grades (r = –.08, p < .05),

November/December 2006 [Vol. 100 (No. 2)] 87

TABLE 1. Intercorrelations Among Primary Variables for Hispanic and Asian Students

Scale 1 2 3 4 5 6 7 8 9 10 11

1. Reading achievement .06 .06 .04 .23** .12* .06 –.07 –.05 .10 .052. Gender –.08* .02 .04 .14* .03 .12* .04 .18** –.12 .103. Grade in school .17** –.05 –.05 –.12* –.11 –.19** –.13* –.17** –.06 –.17**4. Curiosity –.11** .02 –.19** .61** .71** .57** .45** .56** .33** .44**5. Involvement .01 .09** –.22** .67** .65** .60** .41** .56** .42** .45**6. Challenge –.03 .06 –.24** .71** .68** .52** .46** .50** .45** .50**7. Recognition –.01 .08* –.26** .62** .60** .63** .55** .58** .48** .49**8. Grades –.08* .07 –.08* .52** .48** .48** .55** .49** .32** .56**9. Social –.16** .20** –.26** .63** .64** .65** .66** .53** .30** .45**

10. Competition –.02 –.17** –.17** .47** .44** .47** .54** .41** .42** .39**11. Compliance –.01 .06 –.22** .56** .55** .60** .62** .60** .55** .48**

Note. Correlations for Hispanic students are in the lower left section. The upper right section shows correlations for Asian students.*p < .05. **p < .01.

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 9: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

and social (r = –.16, p < .01). For Asian students, readingachievement was positively correlated with involvement(r = .23, p < .01) and challenge (r = .12, p < .05), with nosignificant negative correlations. For Hispanic students,gender (which should be interpreted as a female indicator)was positively correlated with involvement (r = .09, p <.01), recognition (r = .08, p < .05), and social (r = .20, p< .01), but negatively correlated with competition (r =–.17, p < .01). For Asian students, gender was also posi-tively correlated with involvement (r = .14, p < .05),recognition (r = .12, p < .01), and social (r = .18, p < .01),but with no significant negative correlations. While gradein school for Hispanic students was negatively correlatedwith all scales that included intrinsic and extrinsic moti-vation, school grade for Asian students was negativelycorrelated with five scales: involvement (r = –.12, p <.01), recognition (r = –.19, p < .01), grades (r = –.13, p <.01), social (r = –.17, p < .01), and compliance (r = –.17,p < .01). Thus, students in higher grades tended to scorelower on most MRQ scales.

Table 2 compares descriptive statistics and provides t testand effect sizes for variables between Hispanic and Asianstudents. Results specifying that the difference in readingachievement between Hispanic and Asian students was sta-tistically significant (t[1,030] = –6.14, p < .001) indicatethat Hispanic students had lower reading achievement thandid Asian students. That difference was approximately onehalf of a standard deviation (d = –.51), indicating a mediumeffect size. Mean differences for several scales on the MRQwere also statistically significant. For example, recognition(t[1,030] = 1.96, p < .05) and social (t[1,030] = 2.36, p < .05)were statistically higher for Hispanic students, whereascuriosity (t[1,030] = –2.76, p < .05), challenge (t[1,030] =–2.33, p < .05), and compliance (t[1,030] = –2.86, p < .05)were statistically higher for Asian students. Those findings

emphasize the differences that we found in reading motiva-tion between Hispanic and Asian students. Hence, wedecided to use SEM to further explore the differences, aswell as the relationship, with reading achievement.

Factor Structure of Intrinsic and Extrinsic Motivation

To test whether two distinct dimensions could explainthe eight MRQ scales, we conducted CFAs separately forthe Hispanic and Asian groups. For the Hispanic group, atwo-factor model produced a CFI of .98, which indicated avery good model fit. For the Asian group, a two-factormodel also fit very well, producing a CFI of .95. Hence, theresults of CFAs suggested that two factors sufficientlyexplained the variances and covariances among the eightMRQ scales comprising intrinsic and extrinsic motivationfor Hispanic and Asian groups.

Motivational Predictors of Reading Achievement for Hispanicand Asian Groups

One of the questions that we investigated was howintrinsic motivation and extrinsic motivation relates withreading achievement for Hispanic and Asian students. Toanswer the question, we estimated a series of structuralmodels.

We initially estimated a model that used the entire sam-ple (N = 1,032) to test the theoretical model in Figure 1.(The results of the model are shown in Figure 2.) The firstmodel includes ethnicity (Hispanic and Asian) as a predic-tor, which should be interpreted as the contrast betweenAsian and Hispanic students. In other words, it is a binaryindicator for being Asian (Asian = 1, Hispanic = 0).

The next two SEM models separately tested the theoret-ical model in Figure 1 for Hispanic and Asian students. Fig-

88 The Journal of Educational Research

TABLE 2. Descriptive Statistics and t Tests for All Variables in Hispanic and Asian Groups

Hispanic students Asian students(n = 768) (n = 264)

Variable M SD M SD t d

Reading achievement 4.20 1.66 5.13 2.25 –6.14*** –.51Curiosity 2.88 .70 3.01 .60 –2.76** –.19Involvement 2.75 .73 2.80 .65 –1.01 –.07Challenge 2.74 .74 2.85 .62 –2.33* –.15Recognition 2.70 .77 2.59 .75 1.96* .14Grades 3.27 .67 3.29 .58 –.50 –.03Social 2.20 .74 2.09 .65 2.36* .15Competition 2.66 .81 2.70 .79 –.55 –.05Compliance 3.01 .74 3.15 .65 –2.86** –.19Intrinsic motivation 2.79 .64 2.88 .55 –2.12* –.15Extrinsic motivation 2.77 .59 2.76 .51 .16 .02

Note. d = Cohen’s d.*p < .05. **p < .01. ***p < .001.

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 10: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

ure 3 estimates the relationships displayed in Figure 1 forHispanic students (n = 768), and Figure 4 shows the esti-mates of the relationships displayed in Figure 1 for Asianstudents (n = 264). We dropped ethnicity for both of thesubmodels because it is a constant in each group.

Finally, we estimated a multiple-group SEM (see Figure5), treating the Hispanic and Asian groups as “multiplegroups.” That is, we tested the models in Figures 3 and 4simultaneously for Hispanic and Asian groups. That analy-sis allowed us to test not only the equality of parametersacross ethnicity but also whether there were any significantinteractions between ethnicity and the other variables inFigure 1. (For additional information on multiple-groupSEM, see Bentler [1995]).

Following is a summary of the SEM models that we esti-mated to determine the relationships displayed in the the-oretical model in Figure 1:

1. Estimation of the theoretical model using all students(Figure 2);

2. Estimation of the theoretical model using only Hispanicstudents (Figure 3);

3. Estimation of the theoretical model using only Asianstudents (Figure 4); and

4. Estimation of the theoretical model using a multiple-group SEM in which Hispanic and Asian groups aretreated as multiple groups (Figure 5).

The Structural Model for All Students

Figure 2 shows the results of the SEM estimating the the-oretical model displayed in Figure 1 for all students. Thatmodel produced a CFI of .94, and χ2(46) = 342.50, p <.001, a relatively good model fit for all students. In terms ofthe model parameters, grade in school had a negative directeffect on intrinsic motivation (standardized path coeffi-

November/December 2006 [Vol. 100 (No. 2)] 89

Curiosity

FIGURE 2. Estimation of theoretical model for motivation and reading achievement for allstudents; *p < .05.

GradeIntrinsic

Motivation

ExtrinsicMotivation

ReadingAchievement

Recognition

Involvement

Challenge

Gender(Female = 1,

Male = 0)

Grades

Social

Competition

Compliance

Ethnicity(Asian = 1,

Hispanic = 0)

E1.56

E2.60

E3.54

E4.58

E5.74

E6.63

E7.81

E8.68

D1.97

E9.98

D2.96

.83*

.80*

.84*

–.22*

.11*

–.26*

.06

–.05.11*

.07*

.13*

–.02

.91*.12

–.16

.81*

.68*

.78*

.58*

.73*

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 11: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

cient = –.22) and on extrinsic motivation (standardizedpath coefficient = –.26). That finding suggests that,between grades, a higher grade in middle school predictedlower intrinsic motivation and lower extrinsic motivation.That is, older students had lower motivation to read over-all than did younger students. The model also indicatedthat grade in school had a positive direct effect on readingachievement (standardized path coefficient = .11), indicat-ing that, as expected, students in higher grades scoredslightly higher on the reading achievement test.

The predictor variable gender (female indicator) had noaffect on intrinsic motivation. However, it had a small, pos-itive direct effect on extrinsic motivation (standardizedpath coefficient = .11); girls were significantly more affect-ed by extrinsic motivation than were boys. Gender had asmall effect on reading achievement (standardized path

coefficient = –.05, p > .05), indicating that boys scoredslightly, but not significantly, higher than did girls.

The model also indicated that ethnicity (which shouldbe interpreted as an Asian indicator) had a small but sta-tistically significant positive direct effect on intrinsic moti-vation (standardized path coefficient = .07) but not onextrinsic motivation. Ethnicity also predicted readingachievement (standardized path coefficient = .13), indicat-ing that Asian students performed at a statistically signifi-cant higher level than did Hispanic students on the Gates-MacGinitie test.

Structural Model for Hispanic Group

Figure 3 shows the results of SEM estimating the theo-retical model displayed in Figure 1 for the Hispanic stu-

90 The Journal of Educational Research

Curiosity

FIGURE 3. Estimation of theoretical model for motivation and reading achievement forHispanic students; *p < .05.

GradeIntrinsic

Motivation

ExtrinsicMotivation

ReadingAchievement

Recognition

Involvement

Challenge

Grades

Social

Competition

Compliance

Gender

E1.56

E2.59

E3.53

E4.57

E5.72

E6.59

E7.77

E8.66

D1.97

E9.98

D2.95

.83*

.81*

.85*

–.26*

.15*

–.30*

.05*.04

–.07

.92*

.10

–.12

.82*

.72*

.78*

.62*

.75*

.08*

–.17*

.13*

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 12: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

dents. For that group, SEM produced a CFI of .97, andχ2(34, N = 768) = 147.71, p < .001, indicating that thestructural model was a relatively good fit for the Hispanicgroup. Again, grade in school had a negative direct effecton intrinsic motivation (standardized path coefficient =–.26) and extrinsic motivation (standardized path coeffi-cient = –.30), indicating that, between grades, a highergrade in middle school predicted lower intrinsic motiva-tion and extrinsic motivation. The model also indicatedthat school grade has a positive direct effect on readingachievement (standardized path coefficient = .15), demon-strating that students in higher grades scored slightly high-er on the reading achievement test. Grade in school alsohad a direct effect on the individual scale of grades con-tributing to the extrinsic composite (standardized pathcoefficient = .14). That direct effect suggests that theimportance of grades increases for students in higher acad-

emic grades even after adjusting for the effects of intrinsicand extrinsic motivation.

Gender had several direct effects in the Hispanic model.It had a small, positive direct effect on extrinsic motivation(standardized path coefficient = .08) but no effect on intrin-sic motivation. Furthermore, gender had positive directeffects on the involvement dimension of motivation (stan-dardized path coefficient = .05) and the social dimension(standardized path coefficient = .14) but had a negativedirect effect on competition (standardized path coefficient= –.17). Hispanic girls were less competitive and moresocial than were Hispanic boys, even after adjusting for theeffects of extrinsic motivation. Hispanic girls were alsoslightly more involved than were Hispanic boys after adjust-ing for the effect of intrinsic motivation. Thus, it appearsthat Hispanic girls were more likely to have positive effectson involvement in reading and sensitivity to social motiva-

November/December 2006 [Vol. 100 (No. 2)] 91

Curiosity

FIGURE 4. Estimation of theoretical model for motivation and reading achievement for Asianstudents; *p < .05.

GradeIntrinsic

Motivation

ExtrinsicMotivation

ReadingAchievement

Recognition

Involvement

Challenge

Grades

Social

Competition

Compliance

Gender

E1.59

E2.61

E3.55

E4.61

E5.74

E6.69

E7.82

E8.74

D1.99

E9.96

D2.96

.81*

.78*

.84*

–.12

.02

–.22*

.10*.05

.11

.87*

.55*

–.47*

.80*

.68*

.73*

.56*

.67*

.16*

–.21*

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 13: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

tion to read. To summarize, for Hispanic students, neitherintrinsic motivation nor extrinsic motivation had a statisti-cally significant direct effect on reading achievement.

Structural Model for Asian Group

Figure 4 shows the results of SEM estimating the theoreti-cal model displayed in Figure 1 for Asian students. A two-fac-tor measurement model was again imposed on the structuralmodel of reading achievement for the Asian group. Similarto the combined and Hispanic models, the structural modelproduced a CFI of .94, with χ2(36, N = 264) = 101.58, p <.001, indicating a relatively good fit for the Asian group.

The structural model of the Asian data showed that gradein school had a negative direct effect on extrinsic motivation

(standardized path coefficient = –.22) but had no effect onintrinsic motivation, indicating that a higher grade in mid-dle school predicted lower extrinsic motivation but notintrinsic motivation. For Asian students, a higher grade inschool did not contribute to lower intrinsic motivation, aswas the case for Hispanic students. Similar to the Hispanicgroup, grade in middle school also had a direct effect on themotivational dimension of academic grades (standardizedpath coefficient = .14). However, unlike the Hispanic group,grade in school did not predict reading achievement forAsian students after adjusting for the effects of intrinsic andextrinsic motivation (standardized path coefficient = .02).

Similar to the Hispanic group, gender in this model hadseveral significant direct effects. For example, gender had apositive direct effect on the involvement dimension of

92 The Journal of Educational Research

Curiosity

FIGURE 5. Multiple-group (Hispanic/Asian) estimation of theoretical model for motivationand reading achievement; *p < .05.

GradeIntrinsic

Motivation

ExtrinsicMotivation

ReadingAchievement

Recognition

Involvement

Challenge

Grades

Social

Competition

Compliance

Gender

E1 .57/.53

E2 .66/.57

E3 .55/.49

E6 .61/.63

E7 .78/.80

E8 .68/.67

D1 .97/.97

E9 .98/.98

D2 .96/.95

.81*/.85*

.80*/.81*

.84*/.87*

–.22*/–.23*

.13*/.10*

–.28*/–.29*

.07*/.07*.04/.04

–.06/.06

.90*/.90*

.19/.30*

–.21/–.16

.82*/.80*

.10*/.10*

–.18*/–.17*

.13*/0 0/.14*

.70*/.74*

.61*/.62*

.77*/.78*

.74*/.75*

E4 .58/.60

E5 .74/.68

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 14: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

motivation (standardized path coefficient = .10) and a neg-ative direct effect on competition (standardized path coef-ficient = –.21). Like Hispanic girls, Asian girls were lesscompetitive toward reading, although Asian girls weremore involved with it after adjusting for the effects ofintrinsic and extrinsic motivation. In addition, gender hada moderate, positive direct effect on extrinsic motivation(standardized path coefficient = .16) but did not predictsignificant intrinsic motivation.

For Asian students, intrinsic motivation had a consider-able positive direct effect on reading achievement (stan-dardized path coefficient = .55), whereas extrinsic motiva-tion had a considerable negative direct effect on readingachievement.

Multiple-Group Structural Model: Hispanic and Asian Groups

To assess the statistical equality of the observed parame-ter estimates produced by the structural models for the His-panic group (Figure 3) and Asian group (Figure 4), we esti-mated a multiple-group SEM. That helped us explorefurther our third research question: To what extent do thevariables and their relationships in the structural modelsdiffer between Hispanic and Asian middle school students?

For the initial step in a multiple-group SEM analysis, oneestimates the completely unconstrained model. That analy-sis allows all parameters for each group to be estimated sepa-rately. Our completely unconstrained multiple-group SEMresulted in a CFI of .96, and χ2(70, N = 1,032) = 249.29, p <.001. For the next step in a multiple-group SEM analysis, oneestimates the constrained model, which constrains all com-mon parameters between groups. Our constrained model alsoresulted in a CFI of .96, with χ2(89, N = 1,032) = 280.33, p< .001. Lagrange multipliers from the results of the con-strained model indicated that only two constrained parame-ters could be freed and result in a significant decrease in themodel chi-square—the slope predicting reading achieve-ment from gender and the slope predicting reading achieve-ment from intrinsic motivation. Those results indicated (a)a significant interaction effect of gender and ethnicity onreading achievement and (b) a significant interaction effectof intrinsic motivation and ethnicity on reading achieve-ment. After freeing the two parameters, the resulting modelproduced a CFI of .96, and χ2(87, N = 1,032) = 273.12, p <.001. The chi-square difference test from the constrainedmodel produced χ2(2, N = 1,032) = 6.44, p = .04, which wassignificant at the .05 level. Figure 5 shows the results of thefinal multiple-group SEM. All other parameters were con-strained between groups in the constrained and final models,except for those parameters that were set to zero in the Asiangroup but were estimated in the Hispanic group.

Figure 5 shows similar effects, as well as differences, forHispanic and Asian students. For example, the variablegrade had a similar negative effect on intrinsic motivationand extrinsic motivation scales for Hispanic and Asian stu-

dents. Furthermore, gender had a similar small effect onintrinsic motivation (standardized path coefficient = .04, p >.05) for both groups and on extrinsic motivation (standard-ized path coefficient = .10, p < .05) for both groups. Thatsuggested that no statistically significant interactionsoccurred for (a) grade and ethnicity and (b) gender and eth-nicity, in terms of the relationship with reading motivation.

There were some differences in parameter estimatesbetween the Hispanic and Asian groups. For example, theeffect of gender on reading achievement (after adjusting forthe indirect effects of intrinsic and extrinsic motivation)was negative for Hispanic students (standardized path coef-ficient = –.06, p > .05). That indicated that Hispanic boysscored slightly higher than did Hispanic girls on readingachievement. The difference, however, was not statistical-ly significant at the .05 level. Conversely, Asian girls scoredslightly higher than did Asian boys (standardized path coef-ficient = .06, p > .05), but this difference also was not sta-tistically significant at the .05 level. Thus, for both groups,the gender gap in reading seemed to be small and not sta-tistically significant from zero after adjusting for the effectsof intrinsic and extrinsic motivation.

One of the most significant differences in parameterestimates observed in Figure 5 was the differential effect ofintrinsic motivation on reading achievement betweenHispanic and Asian students. The effect of intrinsic moti-vation on reading achievement for Hispanic students wasconsiderably smaller (standardized path coefficient = .19,p > .05) than that for Asian students (standardized pathcoefficient = .30, p < .05). The effect of intrinsic motiva-tion on reading achievement was not statistically differentfrom zero for the Hispanic group. This suggested thatintrinsic motivation had a greater positive impact on read-ing achievement for Asian students than for Hispanic stu-dents. The effect of extrinsic motivation on reading moti-vation was also slightly different for Hispanic students(standardized path coefficient = –.21, p > .05) than forAsian students (standardized path coefficient = –.16, p >.05). Both of those effects, however, were not statisticallysignificant at the .05 level.

Changes in Intrinsic and Extrinsic Motivation Over Time

Structural equation models estimated in Figures 2through 5 identified similarities and differences in terms ofintrinsic and extrinsic motivation and the relationshipwith reading motivation for Hispanic and Asian groups.However, changes in intrinsic and extrinsic motivationover time were also an important aspect of our investiga-tion. McKenna and colleagues (1995) found that attitudetoward recreational and academic reading deterioratedthrough Grades 1 through 6. Because we gathered MRQdata over the fall of 2 years, analysis of that data coulddetermine whether intrinsic and extrinsic motivation forreading changed as students moved from Grade 6 to 7 andfrom Grade 7 to 8. We examined possible intrinsic and

November/December 2006 [Vol. 100 (No. 2)] 93

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 15: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

extrinsic motivational changes in the sample of students forwhom we had data at the two points in time and for possi-ble changes across ethnicities, school grade (6 to 7 and 7 to8), and gender.

First, we explored possible changes in intrinsic and extrin-sic motivation from the fall of Year 1 (Time 1) to the fall ofYear 2 (Time 2) for the combined sample of Hispanic andAsian middle school students (N = 470). Table 3 shows thatfrom Time 1 to Time 2 mean scores on intrinsic motivationdeclined from 2.91 to 2.72, a decrease that was significant atthe .001 level, t(469) = –5.13. Meanwhile, mean scores onthe extrinsic composite fell from 2.77 to 2.65, a drop that wasalso significant at the .001 level, t(469) = –3.26. Thus, thedecline in intrinsic and extrinsic motivation was statisticallysignificant as the students progressed from Time 1 to Time 2,regardless of grade in school or gender.

Second, we disaggregated the sample into Hispanic andAsian groups to determine whether both groups experi-enced significant changes in intrinsic and extrinsic moti-vation over time (see Table 4). For Hispanic students (n =349), the mean score on the intrinsic composite from thefall of Year 1 (Time 1) to the fall of Year 2 (Time 2) droppedfrom 2.89 to 2.69, a decline that was significant at the .001level, t(348) = –4.58. The mean score for the extrinsiccomposite also dropped from 2.77 to 2.66 for Hispanic stu-dents, a decline significant at the .01 level, t(348) = –2.75.For Asian students (n = 121), the mean score on the intrin-sic composite from Time 1 to Time 2 declined from 2.96 to2.79, a drop significant at the .01 level, t(120) = –2.32.However, the drop in extrinsic motivation from Time 1 toTime 2 for the Asian group from 2.75 to 2.62, while numer-ically larger than the drop in extrinsic motivation fromTime 1 to Time 2 for the Hispanic group, was not statisti-cally significant at the .05 level, t(120) = –1.73. Overall,the data indicated that intrinsic motivation in both ethnicgroups declined significantly over time, as cohorts pro-gressed from Grade 6 to 7 and Grade 7 to 8.

Third, we examined changes in the intrinsic and extrinsiccomposites among a cohort of students (n = 265) movingfrom Grade 6 at Time 1 to Grade 7 at Time 2 and a cohort (n

= 205) moving from Grade 7 at Time 1 to Grade 8 at Time 2(see Table 5). For students moving from Grade 6 to Grade 7,mean scores on intrinsic motivation dropped from 3.01 atTime 1 to 2.76 at Time 2, a decline significant at the .001level, t(264) = –4.77. Meanwhile, means on extrinsic moti-vation for this cohort fell from 2.82 to 2.70, a decline signif-icant at the .01 level, t(264) = –2.47. For students movingfrom Grade 7 to Grade 8, mean scores on intrinsic motiva-tion dropped from 2.77 at Time 1 to 2.66 at Time 2, resultingin a difference statistically significant at the .05 level, t(204)= –2.24. Extrinsic motivation also declined for this cohortfrom 2.69 to 2.59, a drop that was also significant at the .05level, t(204) = –2.13. The results leave little doubt about thecontinuing decline in motivation for reading as these stu-dents progressed from Grade 6 to 7 and from Grade 7 to 8.

94 The Journal of Educational Research

TABLE 3. Intrinsic and Extrinsic Motivation Change Over Time for Whole Sample (N = 470)

Time 1 Time 2M SD α M SD α t

Intrinsic motivation2.91 .60 .85 2.72 .62 .86 –5.13*

Extrinsic motivation2.77 .55 .82 2.65 .57 .83 –3.26*

*p < .001.

TABLE 4. Change in Intrinsic and Extrinsic Motivationfor Hispanic and Asian Groups Over Time

Time 1 Time 2M SD M SD t

Hispanic students (n = 349)

Intrinsic motivation2.89 .61 2.69 .64 –4.58***

Extrinsic motivation2.77 .55 2.66 .58 –2.75**

Asian students (n = 121)

Intrinsic motivation2.96 .59 2.79 .57 –2.32*

Extrinsic motivation2.75 .54 2.62 .56 –1.73

*p < .05. **p < .01. ***p < .001.

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 16: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

Finally, we explored intrinsic and extrinsic motivation bygender from Time 1 to Time 2 to determine whether boysand girls registered significant changes (see Table 6). For themale cohort (n = 236), the means for intrinsic motivationfrom Time 1 to Time 2 decreased from 2.88 to 2.65, t(235)= –4.52, p < .001; for extrinsic motivation, the meansdeclined from 2.74 to 2.62, t(235) = –2.37, p < .05. For thefemale cohort (n = 234), the means for intrinsic motivationfrom Time 1 to Time 2 decreased from 2.93 to 2.78, t(233)= –2.78, p < .01; for extrinsic motivation, they dropped from2.80 to 2.69, t(233) = –2.23, p < .05. All of the decliningscores in intrinsic and extrinsic motivation were statistical-ly significant at the .05 level. In summary, those results sug-gest that boys and girls had statistically significant declinesin intrinsic and extrinsic motivation over time.

Discussion

The purposes for this study were (a) to test a theoreticalmodel that included intrinsic and extrinsic motivation,reading achievement, grade in middle school, gender, andethnicity and (b) to clarify previous empirical and theoret-ical work on motivation for reading. The theoretical modelthat we tested arose from the results of prior research on

motivation, especially motivation for reading. The sourcesfor the model came, in part, from research exploring theviability of Self-Determination Theory (SDT). Deci, Ryan,and associates (Deci & Ryan, 1985, 2002; Ryan & Deci,2000) found that intrinsic motivation had a central role inenacting an individual’s progress toward self-regulatedlearning and self-determination. According to SDT, intrin-sically motivated individuals have their locus of controlinternalized and pursue learning because it is personallymeaningful and fulfilling. However, researchers exploringSDT have found that, with few exceptions, as studentsmove through the grades in school, their intrinsic motiva-tion diminishes.

In 1995, McKenna and colleagues (a) developed a modelof motivation for reading, (b) suggested that predictionsabout reading attitudes could be inferred from their model,and (c) found in an empirical study that young students’motivation for reading declined as they moved through theelementary school grades. They also found that (a) leastable readers’ attitudes declined most sharply, (b) girls hadmore positive attitudes than did boys, and (c) ethnicity hadno significant effects on reading attitudes. One componentof the McKenna and colleagues model that may haveexplanatory value in relation to our findings is that of sub-

November/December 2006 [Vol. 100 (No. 2)] 95

TABLE 5. Intrinsic and Extrinsic Motivation AcrossGrades at Time 1 and 2 for All Students

Motivation Students t

Grade Six to Grade Seven Cohort

Intrinsic Grade Six Grade Sevenat Time 1 at Time 2(n = 265) (n = 265)

M 3.01 2.76 –4.77***SD .62 .63

Extrinsic Grade Six Grade Sevenat Time 1 at Time 2(n = 265) (n = 265)

M 2.82 2.70 –2.47***SD .54 .61

Grade Seven to Grade Eight Cohort

Intrinsic Grade Seven Grade Eightat Time 1 at Time 2 (n = 205) (n = 205)

M 2.77 2.66 –2.24*SD .55 .61

Extrinsic Grade Seven Grade Eightat Time 1 at Time 2 (n = 205) (n = 205)

M 2.69 2.59 –2.13*SD .54 .52

*p < .05. **p < .01. ***p < .001.

TABLE 6. Intrinsic and Extrinsic Motivation at Time 1and 2 for Boys and Girls

Motivation Students t

Boys

Intrinsic Grade Six Grade Sevenat Time 1 at Time 2 (n = 236) (n = 236)

M 2.88 2.65 –4.52***SD .61 .60

Extrinsic Grade Six Grade Sevenat Time 1 at Time 2 (n = 265) (n = 265)

M 2.74 2.62 –2.37*SD .55 .56

Girls

Intrinsic Grade Seven Grade Eightat Time 1 at Time 2 (n = 234) (n = 234)

M 2.93 2.78 –2.78**SD .50 .64

Extrinsic Grade Seven Grade Eightat Time 1 at Time 2 (n = 234) (n = 234)

M 2.80 2.69 –2.23*SD .54 .59

*p < .05. **p < .01. ***p < .001.

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 17: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

jective norms, which are contingent on variables like cul-tural identity, identification with that culture, and a rangeof expectations, including those that others impose.

Although previously explained research and theoryinclude most of the key variables in our theoretical model,researchers who developed and applied the MRQ also con-tributed significantly. Those contributions can be foundspecifically in the MRQ scales that compose intrinsic andextrinsic motivation in our theoretical model. Further-more, researchers have used the MRQ to explore grade inschool, gender, ethnicity, intrinsic and extrinsic motiva-tion, and reading achievement. We reported findings relat-ed to those variables in our introduction.

Many critical gaps in our knowledge about motivationfor reading exist; we addressed some of them. How dointrinsic and extrinsic motivation relate to the readingachievement of urban minority middle school students?What relationships exist among gender, grade, intrinsicand extrinsic motivation to read, and reading achieve-ment? How do grade in school, gender, motivation, andreading achievement differ across Hispanic and Asian eth-nicities in an urban middle school? Over time, how doesmiddle school students’ intrinsic and extrinsic motivationchange across ethnicities, school grade, and gender?

Intrinsic and Extrinsic Motivation, Grade, Gender, andReading Achievement

We found that the relation between intrinsic motiva-tion, extrinsic motivation, and reading achievement wasstronger for Asian students than for Hispanic students.After controlling for other variables in the theoreticalmodel for Asian students, intrinsic motivation was posi-tively related to reading achievement at a statistically sig-nificant level, whereas extrinsic motivation was negativelyrelated. However, in the theoretical model for Hispanicstudents, neither intrinsic motivation nor extrinsic motiva-tion had a direct effect on reading achievement that wasstatistically significant. Furthermore, after conducting amultiple-group SEM to examine the relations between thetwo groups, we found that for the Asian group, only intrin-sic motivation had a statistically significant positive effecton reading achievement. The effect of extrinsic motivationfor the Hispanic and Asian groups was negative but did notreach a level of statistical significance for either ethnicity.

The results of the SEM for all students were similar to theestimates for the multiple-group SEM. Mean readingachievement for Asian students was approximately one halfa standard deviation (a full year) greater than for Hispanicstudents, as measured by the Gates-MacGinitie test (seeTable 2). Moreover, intrinsic motivation for Asian studentswas also higher at a .05 level of statistical significance.There was, however, no significant difference in extrinsicmotivation between the Hispanic and Asian groups.

The results of the multiple-group SEM show that gradein school for Hispanic and Asian students had a significant

negative direct effect on intrinsic motivation and extrinsicmotivation, indicating that at one point in time being in ahigher middle school grade contributed to lower intrinsicand extrinsic motivation. As expected, however, grade inschool at one point in time had a positive direct effect onreading achievement.

One of our controlling interests was what becomes of stu-dents’ attitudes toward reading during their progress throughmiddle school. McKenna and colleagues (1995) found thatpositive attitudes toward reading recreational and academictexts decrease from Grades 1–6. We found a similar pattern ofdecreasing motivation for reading as students in our samplemoved through their middle school years. For all the partici-pants, intrinsic and extrinsic motivation declined significant-ly as students moved from Grade 6 at Time 1 to Grade 7 atTime 2, and as they moved from Grade 7 at Time 1 to Grade8 at Time 2. Whereas intrinsic and extrinsic motivation ofthe Hispanic group diminished significantly from Time 1 toTime 2, only intrinsic motivation of the Asian group declinedsignificantly from Time 1 to Time 2. For Asian students,extrinsic motivation did not decline to a level that reachedsignificance. Gender did not significantly affect that patternof deterioration; male and female cohorts declined signifi-cantly for both intrinsic and extrinsic motivation.

Following multiple-group modeling, we found that gen-der had similar direct effects across both ethnicities. ForHispanic and Asian girls, we found a positive direct effecton extrinsic motivation, and on involvement—a compo-nent of intrinsic motivation. For girls of both ethnicities,we found a negative direct effect on competition, a compo-nent of the extrinsic composite. Being a Hispanic girl hada positive direct effect on the social scale, a dimension ofextrinsic motivation. Gender had no direct effects on theintrinsic motivation composite or on reading achievement.Overall, girls rated themselves higher than did boys inextrinsic motivation, suggesting that dimensions of theextrinsic composite, such as recognition, grades, social, andcompliance, contribute to the motivation of girls to agreater degree than to that of boys, who appear to be sig-nificantly more motivated by competition than are girls.

Effects of Intrinsic Motivation on Reading Achievement

As in Wang and Guthrie’s (2004) study of motivationand culture, intrinsic motivation in this study includedthree components: curiosity, involvement, and preferencefor challenge. The authors found that intrinsic motivationpositively related to and predicted text comprehensionequally for their Chinese and American groups. We foundthat intrinsic motivation positively related to and predict-ed reading achievement for Asian students but not His-panic students. The foundation of intrinsic motivation,according to Deci and Ryan’s (1985) Cognitive EvaluationTheory, contains innate psychological needs for compe-tence and self-determination. We suspect that more Asianstudents than Hispanic students were able to pursue and

96 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 18: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

fulfill those needs to expand competence and self-determi-nation. Students engaged in texts that arouse their curiosi-ty, heighten their involvement, and present them withchallenges are likely to build competence and self-determi-nation by using their knowledge, skills, and strategies togain a deeper understanding of books and topics.

The extrinsic composite included recognition, grades,social, competition, and compliance. We anticipated thatextrinsically motivated students read to (a) obtain goodgrades or to improve their grades; (b) gain recognition fromparents, teachers, or schoolmates; (c) experience forms ofsocial connection with family or friends involving reading;(d) show that they read better, receive more correctanswers, or finish reading faster than friends and peers; and(e) comply with a teacher’s reading assignments in a time-ly manner. The extrinsic composite had a significant directnegative affect on reading achievement for the Asian groupin our study, whereas the effect for the Hispanic group wasnot significant at the .05 level. Thus, we do not claim thatextrinsic motivation is generally counterproductive toreading achievement for all students.

Culture and Reading Achievement

The Asian group in our study demonstrated a directeffect of intrinsic motivation on reading achievement (seeFigures 2 and 5) that echoed the finding by Wang andGuthrie (2004) that intrinsic motivation supported textcomprehension. However, the intrinsic and extrinsic com-posites of the Hispanic group did not have a significantdirect effect, positive or negative, on reading achievement.Nor did the estimated model for the Hispanic group (seeFigure 3) demonstrate any effects of either intrinsic orextrinsic motivation. Those differences that may have cul-tural roots warrant further exploration.

The Asian students may have internalized values andbeliefs related to schooling that parallel those of students inWang and Guthrie’s (2004) Chinese and U.S. samples.Wang and Guthrie inferred that those values and beliefswere likely to facilitate engagement of the Chinese andU.S. students in their study with the affective and cogni-tive challenges of reading and understanding texts. We sus-pect that the Asian urban middle school students in oursample were influenced by similar internalized cultural val-ues and beliefs about schooling. In their review of researchon Asian cultural influences, Wang and Guthrie noted thatmiddle-income Chinese students emphasize academic suc-cess more so than do middle-income American studentswith whom they compared the Chinese students’ motiva-tion to read.

Wang and Guthrie (2004) cited research supporting theChinese emphasis on education and academic success as ameans of promoting one’s self and achieving social status.Those internalized cultural values and beliefs may con-tribute to the “subjective norms” influencing reading moti-vation in the reading-attitude model developed by McKen-

na and colleagues (1995). If that were the case in this study,we would have expected to find that the Asian studentswere more exposed to cultural influences, such as havingopportunities to explore and to discover topics of interest,which nurtured the evolution of intrinsic motivation, thedevelopment of individual competence, and self-determi-nation. The Asian students may also have valued externalregulations or internalized regulations, such as high expec-tations for grades to gratify their parents, so that theyembraced the regulations and expectations as arising fromtheir core values (Ryan & Deci, 2000).

Other factors may contribute to the apparent absence ofeffects of intrinsic motivation in the Hispanic group andthe apparent differences in the impact of intrinsic andextrinsic motivation. The differences in motivationbetween the Asian and Hispanic student samples could beaffected by the different orientations that various culturalgroups have toward school.

Research conducted and theory developed by education-ally oriented cultural anthropologists may offer someinsight into the differences in motivation between theAsian and Hispanic students in our study. For example, theexplanation that Ogbu (1983) advocated became embod-ied in his cultural-ecological theory of minority school per-formance. Ogbu’s cultural-ecological theory has two parts:(a) the manner in which minorities are treated and reward-ed in education settings and (b) the way that minoritiesperceive and respond to their treatment in school settings.He also pointed out that the manner in which a groupbecame a minority affected that minority’s responses.

Ogbu (1983, 1991) classified minorities into voluntaryand involuntary groups; voluntary minorities are those whoimmigrated to America willingly, usually to improve theiropportunities and those of their children. Examples of vol-untary minorities include Korean, Japanese, and Mexicanimmigrants. In contrast, involuntary minorities have unwill-ingly become part of American society. Having beenenslaved, colonized, or conquered, involuntary minoritiesusually view being in America as a condition forced onthem. Examples of involuntary minorities include AfricanAmericans, Native Americans, and early Mexican Ameri-cans who were conquered.

Ogbu (1983, 1991; Ogbu & Simons, 1998) proposed thatvoluntary and involuntary minorities develop and internal-ize cultural models of American society that enable groupmembers to interpret their world and respond to it. Volun-tary minorities perceive American society, its culture, andits schools in ways that align far more closely with a visionof self-determination than do the perceptions of involun-tary minorities. Voluntary minorities, such as the Asiansample in our study, frame their situation in America as onethat promotes self-realization through the pursuit of oppor-tunity in a new land. The belief of voluntary minoritiesthat hard work and education will enable them to achievetheir vision of a self-determined future provides them witha significant degree of self-regulation and internal locus of

November/December 2006 [Vol. 100 (No. 2)] 97

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 19: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

control. Voluntary minorities tend to be optimistic thatAmerican institutions, like schools, will facilitate their pur-suit of new lives of promise. Moreover, they believe thatthe new culture and language they are learning will aug-ment their identity rather than subvert the culture and lan-guage of their heritage. The internalized “cultural model”of American society that voluntary minorities constructtends to place them at the center of their locus of control.The compatibility between their view of America and theirself-determined futures provides a social situation in whichintrinsic motivation, including the intrinsic motivation oftheir children, can grow and propel them toward their goalsand aspirations.

Conversely, for involuntary minorities, the path toadopting America’s culture and language is likely to befilled with far more conflict, ambivalence, and hesitation(Ogbu, 1991; Ogbu & Simons, 1998). Many features intheir cultural model of America result in placing their locusof control outside themselves. Involuntary minorities, per-haps like the families of some of the Hispanics in our study,may frame their situation in America as one that is far lesspromising in social and economic terms. Hard work andeducation may not enable them to achieve their vision of aself-determined future. Involuntary minorities tend to bemore pessimistic about American institutions and theirability to facilitate the realization of new lives. They mayset their economic sights lower than voluntary minorities,see “job ceilings,” and meet job expectations working inconstruction or department stores (Matute-Bianchi, 1991).For some immigrants, American institutions, like theimmigration services, may inhibit their dreams. Moreover,the new culture and language they are expected to absorbmay threaten their heritage and identity. Most, if not all, ofthose forces are outside themselves. Intrinsic motivationwould be difficult to foster in institutions, like schools, thatendanger the realization of current and future identities.Thus, in-school literacies, including the acquisition of Eng-lish, may not always be perceived as funds of knowledgethat will enable them, without conflict, to pursue theirvision of a self-determined future.

Although the cultural-ecological explanation of minor-ity school success and failure has explanatory appeal,research indicates that school contexts exercise a powerfulinfluence on minority student performance (Conchas,2001, 2006; Conchas & Clark, 2002; Kozol, 1991). Thevoluntary/involuntary or immigrant/nonimmigrant cate-gories may simplify a motivational picture of minorityschooling that is, when examined closely, far more com-plex than these categories imply. Sociocultural processeswithin schools, rather than cultural-ecological processes,may contribute to a more accurate explanation of differ-ences in minority student performance. Whereas someschool cultures and within-school academies or smalllearning communities (SLC) may not optimally promoteinclusion and provide support for minority students, othersare highly successful in developing and sustaining social

support systems for minority youth, including Asian andHispanic students. For example, low-achieving Latino stu-dents are often invisible in some school cultures and learn-ing communities, but in others they receive ample supportthrough programs like AVID (Advancement Via Individ-ual Determination). The sociocultural processes withinAVID, SLCs, or other academy-like school structures canpromote positive perspectives of social mobility andenhance students’ engagement with literacy and learning.

Implications for Research

Although Ogbu and his associates (Ogbu, 1981, 1983,1991; Ogbu & Simons, 1998) have provided explanationsthat contribute to our understanding of differences inresponse to American schools by voluntary and involun-tary minorities, gaps in our understanding of cultures andschooling remain and warrant further exploration. Where-as the importance of intrinsic motivation to schoolingappears to be well established, the role that cultures play inthe genesis of that motivation is far less clear. This studyand many others have shown that significant relations existbetween intrinsic motivation and school achievement inmultiple forms, including that of reading achievement. Ofsignificant value would be knowledge about the originationand growth of intrinsic motivation in individuals. Howdoes intrinsic motivation for reading evolve in children,and how can it be maintained or promoted? What roles doparents and the culture with which they identify play infostering intrinsic motivation? What roles do mothers,fathers, siblings, and the extended family play? Do familiestransmit cultural expectations that establish an internallocus of control which, when carried into school settings,enables students to become involved in learning, to acti-vate curiosity, and to seek challenges? Do patterns of cul-tural transmission and influence vary across cultures oreven among subgroups within cultures? How do the trans-fer of these cultural values, beliefs, and expectations differacross gender?

The contributions of learning experiences in school set-tings to the motivational orientation of middle school stu-dents also warrants future investigation. Although theeffects on reading motivation of a wide range of within-school learning experiences were not a major focus of thisstudy, many factors that teachers can influence contributeto students’ engagement in reading (Ruddell & Unrau,1997; Unrau, 2004). The factors include teacher engage-ment, achievement-related instruction, classroom commu-nity, assessment and reward systems, and autonomy sup-port. Ruddell (2004) conducted theoretical- andresearch-based studies that provide a portrait of influentialteachers who heightened their students’ motivation to readby arousing their students’ curiosity and by providing themwith opportunities to engage in the challenges of problemsolving. However, much research needs to be performed onthe impact of influential, engaged teachers. What goals,

98 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 20: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

beliefs, and practices of influential teachers heighten stu-dents’ motivation for reading, especially during episodes ofmeaning negotiation in classroom communities?

Students who have acquired reading strategies and back-ground knowledge that enable them to make connectionswith books and to read grade-level texts proficiently are farmore likely to enjoy a stronger sense of self-efficacy andlower amounts of frustration when engaging in school read-ing assignments (Duke & Pearson, 2002). However, ourknowledge of the impact on motivation for reading a rangeof literacy-promoting programs, including those with directinstruction, extensive technology, or a master-apprenticeframework, implemented at the elementary and middleschool levels is limited. During an era in which more script-ed, phonics-based programs are being used not only in thelower grades but also in intervention programs for strugglingreaders in the higher grades, educators and researcherswould benefit from knowing what influence these programshave on students’ long-term motivation for reading.

Educators know that third- and fifth-grade students’ useof strategies, reading comprehension, intrinsic motivation,and reading self-efficacy increase when their students’learning experience includes Concept Oriented ReadingInstruction (CORI; Guthrie et al., 1998; Wigfield, Guthrie,Tonks, & Perencevich, 2004). CORI is an approach thatcombines several sequenced features to enhance students’intrinsic motivation: (a) hands-on science activities tospark interest, (b) question generation following thoseactivities, (c) connection of students’ interests and ques-tions to relevant books, and (d) meaningful choices forindependent learning to promote autonomy. Wouldinstruction modeled on CORI principles and implementedthroughout content courses during students’ elementaryand middle school years heighten intrinsic motivation,reading engagement, strategy use, and learning from texts?

Other approaches that enhance intrinsic motivation andmerit further investigation involve autonomy support(Assor, Kaplan, & Roth, 2002). Also, Reeve, Jang, Carrell,Jeon, & Barch (2004) trained high school teachers acrosscontent areas to nurture student interest, provide rationalesfor assignments, and use noncontrolling language. Teacherswho were trained displayed more autonomy support whenworking with students than did those who were nottrained. Furthermore, the more teachers used autonomysupport during instruction, the more their students wereengaged (Reeve & Jang, 2006). Could literacy programs atthe elementary and middle school levels that emphasizeautonomy support and a wide range of choices in learningopportunities promote intrinsic motivation and enable stu-dents to engage more deeply in literacy later in theirschooling?

The percentage of first-generation 16- to 24-year-oldHispanics who were high school dropouts in 2000 wasabout three times as high as that for non-Hispanics; for sec-ond-generation 16- to 24-year-olds, it was about twice ashigh (Kaufman, Alt, & Chapman, 2001). Many of those

students begin the process of dropping out while in middleschool, as they become increasingly disengaged from schooland schooling. Ogbu (1983, 1991) described some of themany reasons for disengagement. However, the loss of edu-cational opportunities to those students cannot be calcu-lated. What motivational profiles related to literacy arebehind their decisions to quit school? Knowing that moti-vation for school declines precipitously for many Hispanicstudents during their middle school years, what can bedone to improve their prospects and preparation for highschool? Work on adolescent literacy (Alvermann, 2001)has focused on adolescent self-efficacy and engagement, aswell as on culturally responsive instruction. In addition,several federally funded programs, including AVID andGEAR UP (Gaining Early Awareness and Readiness forUndergraduate Programs), address the academic needs andmotivation of minority students in urban settings. Thoseinitiatives would benefit from our learning more aboutmotivational development, its roots, and threats to itsvitality in populations of students who tend to be more atrisk of dropping out of school.

To proponents of self-determination theory, the drivetoward more rigorous high-stakes testing programs are like-ly to inflict enormous harm on students’ motivation, includ-ing the intrinsic motivation of those students who are curi-ous, involved, and ready for authentic learning challenges(Ryan & La Guardia, 1999). Many literacy educators arealso concerned about the potential effects of those programson literacy motivation. Students at risk may be put atgreater risk of disengagement from school and from literacydevelopment because of high-stakes testing programs. Thatfinding is particularly ironic because many state content andperformance standard programs were adopted to addressgaps in the performance of students from diverse ethnicities.Self-determination theorists and researchers argue thatreward- and punishment-oriented approaches that pressurestudents cannot succeed. Educators, researchers, and policymakers need to know more about the effects of statewidehigh-stakes assessment programs on the reading motivationof voluntary and involuntary minority students to morewisely assess those programs. Do cultural factors exist with-in immigrant and nonimmigrant groups that affect respons-es, positive and negative, to the programs? If so, what arethose factors and how do they operate?

Limitations of the Research

Although we did not include ESL students in this study,some students not classified as ESL may still have had trou-ble comprehending questions because of limited readingskills. That limitation could apply to the Hispanic andAsian students in this study.

Another limitation in terms of external validity ofresults is that our data were not the result of random sam-pling, and, thus, the results should not be generalized to allschools. However, it is reasonable to suggest that the results

November/December 2006 [Vol. 100 (No. 2)] 99

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 21: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

could generalize to other urban schools with high percent-ages of low-SES, Hispanic and Asian students. Further-more, schools may include different ethnic or cultural sub-groups that could present various patterns of motivation.When Matute-Bianchi (1991) examined the subgroupsthat made up the Hispanic population at a particular highschool, she discovered a range of subgroups, from recentimmigrants to “Cholos,” whose profiles varied and whoseengagement with school differed. Other studies haveshown links between students’ sense of Hispanic subgroupidentity, such as that of Chicano students, their broaderidentity as individuals, and their degree of academicengagement (Pizarro, 2005). Nevertheless, the resultsreported from our study of Asian and Hispanic studentspaint a picture of differences in motivation and their alter-ation in an urban middle school.

A further limitation of our results is that we did notdirectly address the range of learning experiences thatcould have influenced participants’ motivation for readingprior to entering middle school. Several of those learningexperiences could have affected student performance andengagement with reading. For example, the degree towhich the teacher was enthusiastic about teaching, theteacher’s pedagogy or motivational orientation, and theoverall culture of the elementary school could have affect-ed student achievement and motivation for reading.

In addition, the relationships estimated by the SEMmodels do not suggest causal links. Because of the complexnature of the variables and the relationships involved withstudent reading motivation and student reading achieve-ment, the relationships estimated by the SEM models donot necessarily imply causality, particularly because socialeconomic status, English-language-learning status, andother social factors were not variables in our study. There-fore, the relationships estimated and tested by the SEMmodels should be used only for comparison purposes.

Summary

Our results suggest that intrinsic motivation is a salientfactor affecting the reading achievement of Asian middleschool students. Intrinsic motivation had a smaller effect,however, on the reading achievement of Hispanic middleschool students. Moreover, intrinsic motivation seemed tohave a positive effect on reading achievement for Hispanicand Asian middle school students, whereas extrinsic moti-vation seemed to have a negative impact on readingachievement. Furthermore, there were few gender differ-ences in terms of the effects of intrinsic and extrinsic moti-vation on reading achievement; however, after adjustingfor the effects of intrinsic and extrinsic motivation, femalestudents (Hispanic and Asian) seemed to be more socialand less competitive than their male counterparts. Thatfinding suggests that there were gender differences in socialmotives beyond those explained by intrinsic and extrinsicmotivation. Finally, there was a statistically significant

decline in intrinsic and extrinsic motivation over time forHispanic and Asian middle school students.

This investigation highlights the importance of discover-ing more about the impact of motivation, especially that ofintrinsic motivation, on the engagement of middle schoolstudents and the relationship of that motivation to theirgrowth in literacy and learning. The apparent worsening ofstudents’ attitudes toward reading from Grades 1–8 furthersuggests that educators face relentless challenges to students’reading engagement. Discovering the factors contributing tostudents’ reading motivation, especially for students in themiddle school years, and the means to deepen engagementin reading and learning could promote the transformation ofmany classrooms and the lives of many students.

REFERENCES

Alvermann, D. E. (2001). Effective literacy instruction for adolescents. Exec-utive summary and paper commissioned by the National Reading Con-ference. Chicago, IL: National Reading Conference.

Assor, A., Kaplan, H., & Roth, G. (2002). Choice is good, but relevanceis excellent: Autonomy-enhancing and suppressing teacher behaviorspredicting students’ engagement in schoolwork. British Journal of Edu-cational Psychology, 72, 261–278.

Bentler, P. M. (1995). EQS structural equations program manual. Encino,CA: Multivariate Software, Inc.

Baker, L., & Wigfield, A. (1999). Dimensions of children’s motivation forreading and their relations to reading activity and reading achievement,Reading Research Quarterly, 34, 452– 477.

Chen, C. S., Stevenson, H. W., Hayward, C., & Burgess, S. (1995). Cul-ture and achievement: Ethnic and cross-cultural differences. In M. L.Maehr & P. R. Pintrich (Eds.), Advances in motivation and achievement(Vol. 9, pp. 119–151). Greenwich, CT: JAI Press.

Conchas, G. (2001). Structuring failure and success: Understanding thevariability in Latino school engagement. Harvard Educational Review,70, 475–504.

Conchas, G. (2006). The color of success: Race and high-achieving urbanyouth. New York: Teachers College Press.

Conchas, G., & Clark, P. A. (2002). Career academies and urban minor-ity school success: Forging optimism despite limited opportunity. Jour-nal of Education for Students Placed at Risk, 7(3), 287–311.

Csikszentmihalyi, M. (1990). Literacy and intrinsic motivation. Daedalus,119(2), 115–140.

Deci, E. L., Koestner, R., & Ryan, R. M. (2001). Extrinsic rewards andintrinsic motivation in education: Reconsidered once again. Review ofEducational Research, 71, 1–27.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determina-tion in human behavior. New York: Plenum Press.

Deci, E. L., & Ryan, R. M. (Eds.). (2002). Handbook of self-determinationresearch. Rochester, NY: University of Rochester Press.

Duke, N., & Pearson, P. D. (2002). Effective practices for developing read-ing comprehension. In S. J. Samuels & A. E. Farstrup (Eds.), Whatresearch says about reading instruction (3rd ed., pp. 205–242). Newark,DE: International Reading Association.

Eccles, J. S., Lord, S., & Buchanan, C. M. (1996). School transitions inearly adolescence: What are we doing to our young children? In J. A.Graber, J. Brooks-Gunn, & A. C. Peterson (Eds.), Transitions throughadolescence: Interpersonal domains and context (pp. 251–284). Hillsdale,NJ: Erlbaum.

Gottfried, A. E., Fleming, J. S., & Gottfried, A. W. (2001). Continuity ofacademic intrinsic motivation from childhood through late adoles-cence: A longitudinal study. Journal of Educational Psychology, 93, 3–13.

Graham, S. (1994). Motivation in African Americans. Review of Educa-tional Research, 64, 55–118.

Guthrie, J. T., McGough, K., & Wigfield, A. (1994). Measuring readingactivity: An inventory (Instructional Resource No. 4). Athens, GA:National Reading Research Center, Universities of Georgia and Mary-land College Park.

Guthrie, J., Van Meter, P., Hancock, G. R., Alao, S., Anderson, E., &

100 The Journal of Educational Research

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014

Page 22: Motivation and Its Relationship With Reading Achievement in an Urban Middle School

McCann, A. (1998). Does concept-oriented reading instructionincrease strategy use and conceptual learning from text? Journal of Edu-cational Psychology, 90, 261–278.

Guthrie, J., Wigfield, A., Metsala, J. L., & Cox, K. E. (1999). Motiva-tional and cognitive predictors of text comprehension and readingamount. Scientific Studies of Reading, 3, 231–256.

Harter, S. (1981). A new self-report scale of intrinsic versus extrinsic ori-entation in the classroom: Motivational and informational compo-nents. Developmental Psychology, 17, 300–312.

Kaufman, P., Alt, M. N., & Chapman, C. D. (2001). Dropout rates in theUnited States: 2000. Washington, DC: National Center for EducationStatistics, U.S. Department of Education. NCES 2002-114.

Kohn, A. (1993). Punished by rewards: The trouble with gold stars, incentiveplans, A’s, and other bribes. New York: Houghton Mifflin.

Kozol, J. (1991). Savage inequalities: Children in American schools. NewYork: HarperCollins.

Lepper, M. R., Corpus, J. H., & Iyengar, S. S. (2005). Intrinsic and extrin-sic motivational orientations in the classroom: Age differences and aca-demic correlates. Journal of Educational Psychology, 97, 184–196.

Lepper, M. R., & Henderlong, J. (2000). Turning “play” into “work” and“work” into “play”: 25 years of research on intrinsic and extrinsic moti-vation. In C. Sansone & J. M. Harackiewicz (Eds.), Intrinsic and extrin-sic motivation: The search for optimal motivation and performance (pp.257–307), San Diego, CA: Academic Press.

Lepper, M. R., Sethi, S., Dialdin, D., & Drake, M. (1997). Intrinsic andextrinsic motivation: A developmental perspective. In S. S. Luthar, J.A. Burack, D. Cicchetti, & J. R. Weisz (Eds.), Developmental psy-chopathology: Perspectives on adjustment, risk, and disorder (pp. 23–50).New York: Cambridge University Press.

Lipson, M. Y., & Wixson, K. K. (1997). Assessment and instruction of read-ing and writing disability: An interactive approach (2nd ed.). New York:Longman.

Liska, A. E. (1984). A critical examination of the causal structure of theFishbein/Ajzen attitude-behavior model. Social Psychology Quarterly,47, 61–74.

MacGinitie, W. H., & MacGinitie, R. K. (1998). Gates-MacGinitie Read-ing Test, Level 5/6 (3rd ed.). Chicago: Riverside.

Matthewson, G. C. (1976). The function of attitude in the readingprocess. In H. Singer & R. B. Ruddell (Eds.), Theoretical models andprocesses of reading (2nd ed., pp. 655–676). Newark, DE: InternationalReading Association.

Matute-Bianchi, M. (1991). Situational ethnicity and patterns of schoolperformance among immigrant and nonimmigrant Mexican-descentstudents. In M. A. Gibson & J. Ogbu (Eds.), Minority status and school-ing: A comparative study of immigrant and involuntary minorities (pp.295–247). New York: Garland.

McKenna, M. C., Kear, D. J., & Ellsworth, R. A. (1995). Children’s atti-tudes toward reading: A national survey. Reading Research Quarterly, 30,934–956.

Ogbu, J. U. (1981). Origins of human competence: A cultural-ecologicalperspective. Child Development, 67, 1–13.

Ogbu, J. (1983). Minority status and schooling in plural societies. Com-parative Education Review, 27(2), 168–190.

Ogbu, J. (1991). Immigrant and involuntary minorities in comparativeperspective. In M. A. Gibson & J. Ogbu (Eds.), Minority status andschooling: A comparative study of immigrant and involuntary minorities (pp.3–33). New York: Garland.

Ogbu, J., & Simons, H. (1998). Voluntary and involuntary minorities: Acultural-ecological theory of school performance with some implica-tions for education. Anthropology & Education Quarterly, 29, 155–188.

Otis, N., Grouzet, F., & Pelletier, L. G. (2005). Latent motivationalchange in an academic setting: A 3-year longitudinal study. Journal ofEducational Psychology, 97(2), 170–183.

Pizzaro, M. (2005). Chicanas and chicanos in school: Racial profiling, identitybattles, and empowerment. Austin, TX: University of Texas Press.

Reeve, J., & Jang, H. (2006). What teachers say and do to support stu-dents’ autonomy during a learning activity. Journal of Educational Psy-chology, 98, 209–218.

Reeve, J., Jang, H., Carrell, D., Jeon, S., & Barch, J. (2004). Enhancingstudents’ engagement by increasing teachers’ autonomy support. Moti-vation and Emotion, 28, 147–169.

Rowling, J. K. (1998). Harry Potter and the sorcerer’s stone. New York:Arthur A. Levine Books.

Ruddell, R. B. (2004). Researching the influential literacy teacher: Char-acteristics, beliefs, strategies, and new research directions. In R. B. Rud-dell & N. J. Unrau (Eds.), Theoretical models and processes of reading (5thed., pp. 979–997). Newark, DE: International Reading Association.

Ruddell, R. B., & Unrau, N. J. (1997). The role of responsive teaching infocusing reader intention and developing reader motivation. In J. T.Guthrie & A. Wigfield (Eds.), Reading engagement: Motivating readersthrough integrated instruction (pp. 102–125). Newark, DE: InternationalReading Association.

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and thefacilitation of intrinsic motivation, social development, and well-being.American Psychologist, 55, 68–78.

Ryan, R. M., & La Guardia, J. G. (1999). Achievement motivation with-in a pressured society: Intrinsic and extrinsic motivations to learn andthe politics of school reform. In T. Urdan (Ed.), Advances in motivationand achievement (Vol. 11, pp. 45–85). Greenwich, CT: JAI Press.

Steinberg, L. (1996). Beyond the classroom: Why school reform has failed andwhat parents need to do. New York: Simon & Schuster.

Stevenson, H. W., Chen, C., & Uttal, D. H. (1990). Beliefs and achieve-ment: A study of black, white, and Hispanic children. Child Develop-ment, 61, 508–523.

Thernstrom, A., & Thernstrom, S. (2003). No excuses: Closing the racialgap in learning. New York: Simon & Shuster.

Unrau, N. J. (2004). Content area reading and writing: Fostering literacies inmiddle and high school cultures. Upper Saddle River, NJ: Pearson/Mer-rill/Prentice-Hall.

Wang, J. H., & Guthrie, J. (2004). Modeling of effects of intrinsic moti-vation, extrinsic motivation, amount of reading, and past readingachievement on text comprehension between U.S. and Chinese stu-dents. Reading Research Quarterly, 39, 162–186.

Watkins, M. W., & Coffey, D.Y. (2004). Reading motivation: Multidi-mensional and indeterminate. Journal of Educational Psychology, 96,110–118.

Wigfield, A., & Guthrie, J. T. (1995). Dimensions of children’s motivationsfor reading: An initial study (Reading Research Report No. 34). Athens,GA: National Reading Research Center, Universities of Georgia andMaryland College Park.

Wigfield, A., & Guthrie, J. T. (1997). Relations of children’s motivationfor reading to the amount and breadth of their reading. Journal of Edu-cational Psychology, 89, 430–432.

Wigfield, A., Guthrie, J. T., & McGough, K. (1996). A questionnaire mea-sure of children’s motivations for reading. (Instructional Resource No. 22).Athens, GA: National Reading Research Center, Universities of Geor-gia and Maryland College Park.

Wigfield, A., Guthrie, J. T., Tonks, S., & Perencevich, K. (2004). Chil-dren’s motivation for reading: Domain specificity and instructionalinfluences. The Journal of Educational Research, 97, 299–309.

November/December 2006 [Vol. 100 (No. 2)] 101

Dow

nloa

ded

by [

130.

108.

121.

217]

at 1

1:01

26

Aug

ust 2

014


Recommended