DOCUMENT RESUME

ED 335 518 CE 058 735

AUTHOR Osman-Gani, A. Ahad M.; And OthersTITLE Working in Groups: Selected Research Abstracts.

Training and Development Research Report.INSTITUTION Ohio State Univ., Columbus. Coll. of Education.PUB DATE Sep 91NOTE 27p.

PUB TYPE Reference Materials - Bibliographies (131)

EDRS PRICE MF01/PCO2 Plus Postage.DESCRIPTORS Abstracts; *Goal Orientation; *Group Behavtor; *Group

Dynamics; *Group Experience; Groups; *JobPerformance; *Productivity

ABSTRACTThis report contains abstracts of 20 selected

research studies related to individuals working as a group to achievea goal. Each abstract consists of these Jomponentl: author(s) orresearcher(s); title; source (journal citation); purpose :A theresearch; participants; method; results; and discussion. In general,the studies show that working in groups can improve the performanceof individuals and the group. The report is intended to help guidehuman resource development professionals in their efforts to designmore productive work systems. ;YLB)

*****************************************************m*****************

Reproductions supplied by EDRS are the test that can be madefrom the original document.

***********************************************************************

Working in Groups: Selected Research

S DIPARTIIIPIT OP EDUCAlleiee Fttutatrimus Romer/. Aral Innrovirt'ATONAL RE SOuRCES ffvFORt.$ATION

CENTE IMOTN% CIOKunsent hdb i)per, *rwpclut pa isfeconoed Me Demon o 019811,1104(1110,1014,MASO

r Minot chomps have been mtOSP anotomPrOtatk#11or cluabiA

KuntS 0' rps. 0. torposs shatA0 1,1,5

mew do not nrIttorEdt4r 0",t40Of RI pofot 41ft CO t r

Abstracts

A. Ahad M. Osman-Grani

Shashi B. Gowda

Patrick E. Smylie

Ronald L Jacobs

-PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY

TO THEINFORMA

Training and Development

Research Report

ATIONAL RESOURCESN CENTER IERICL-

A report completed as part of the Synchonous Work Group

Research and Development Project

September 1991

BEST AVAILABLE

2

Much has been reported in the popular and scholarly

literature about working in groups in organizations. Of all

the efforts to increase productivity and improve quality,

none seems to hold more interest among managers and workers

than this work design. Work groups are an approach to work

in which workers share responsibility and accountability to

complete a ptoduct or service that meets customer

requirements. Use of work groups has resulted in positive

outcomes for organizations and workers. Organizations have

reported fewer defects, lower unit costs, and increased

workforce efficiency. Workers have reported increased

feelings of accomplishment, greater involvement with the

goals of the organization, and increased potential for self-

development.

This report addresses a practical question that is

frequently asked: "What are the critical features of work

groups?" This question was addressed by reviewing the

literature related to individuals working as a group to

achieve a goal. In general, the studies show that working in

groups can improve the performance of individuals and the

group. The report was prepared to help guide human resource

development professions in their efforts to design more

productive work systems.

A. Ahad Osman-GaniShashi B. GowdaPatrick E. SmileyRonald L. Jacobs

13

Erez, M., & Zidon, I. (1984). Effect of goal acceptance onthe relationship of goal difficulty to performance.journal of Applied Psychology, 69(1), 69-78.

Purpose To determine the impact of goal acceptance anddifficulty on group performance.

Participants

Method

Results

Discussion

140 technicians and engineers.

Participants were instructed to complete aperceptual speed test requiring them todetermine how many digits or letters in a rowwere the same as the circled one to the leftof each row. Performance was measured by thenumber of numerals and characters correctlycircled within an allotted time period.

The results were as follows:

Performance increased with goal difficultyin the high and medium-goal acceptancegroups.

Performance decreased as goals became moredifficult in the low-goal acceptance group.

Goal acceptance decreased with goaldifficulty for all groups.

The results demonstrated the relationshipbetween performance level and goal difficultywhen goals are accepted and rejected. Goaldifficulty offers a frame of reference forevaluating how people respond to differentgoal situations.

Latham, Go, & Yukl, G. (1976). Assigned versus participativegoal setting with educated and uneducated wood workers.journal of Applied Psychology, 60(3), 299-312.

Purpose To study the effects of participative andassigned goal setting in work groups.

Participants

Method

Results

Discussion

48 logging crews in the wood industry.

Organization managers were told that a methodhad been found which that allow them toincrease their productivity at no cost to themor their workers. Participative and assignedwork groups were established. Participativework groups asked their employees each week toset a specific production goal in terms ofcunits (100 cubic feet of wood per week) for aeight week period). Assigned work groups weregiven the production goal.

The results were the following:

Performance was higher in the participativegoal-setting condition than in the assignedor "do your best" goal-setting conditions.

Goal difficulty was higher in theparticipative goal-setting condition than inthe assigned goal-setting condition.

Goal achievement occurred more often in theparticipative goal-setting condition than inthe assigned goal-setting condition.

The results suggest that participation affectsperformance when there is the establishment ofand the commitment to specific goals. Thesuperiority of participative goal setting isdue in part to a higher goal being set by thelogging crews. The acceptance of goals andthe motivation to attain those goals isgreater when employees are allowed toparticipate in the goal setting process.

3

Matsui, T., Kakuyama, T., & Uy Onglatco, M. (1987).. Effectsof goals and feedback on performance in groups. journalof Applied Psychology, 72(3), 407-415.

Purpose To determine 1) whether people would performbetter when the performance standard was groupoutput rather than individual output and 2)whether group goal-setting would lead tohigher performance than individual goalsetting when goal acceptance is higher.

Participants In Study 1, 104 college men and women.

Method

In Study 2, 100 college men and women.

In Study 1, team members were told to specifythe score their team would achieve and thendiscuss the score each member would contributeto the group goal on a numerical countingtask.

In Study 2, team members were asked to sittogether on a similar numerical counting task,and were informed of their team's ability.They were told that there would be a triallasting 15 minutes, with the goal ofcompleting 160 rows as a team.

Results The results were the following:

In Study 1, group-goal participantsperformed higher than individual-goalparticipants, who had higher goals. Goalacceptance was higher for group-goalparticipants than for individual-goalparticipants.

In Study 2, although participants workedtogether as teams, individual team-memberdata were gathered. Post-feedbackperformance was higher than pre-feedbackperformance suggesting that task feedbackenhances performance. Higher abilitysubjects maintained their motivationalthough they learned their lower abilitypartners were performing poorly.

4 6

Discussion The results showed that feedback on both groupand individual performance data maximizesefforts. Having members work as teams with aspecific team goal, rather than as individualswith an individual goal, increasesproductivity.

75

Koch, J. (1979). Effects of goal specificity and performancefeedback to work groups on peer leadership, performance,and attitudes. Human Relations, 32(10), 819-840.

Purpose

Participants

Method

Results

Discussion

To examine how goals and feedback in workteams relates to leadership, groupeffectiveness, performance, and attitudes.

78 female piecework operators.

A questionnaire was administered to allparticipants approximately seven months beforegoal setting and feedback programs. Plexiglasdisplay boards were used to show performancedata to five subassembly groups, who had beenidentified by operators in "natural" workunits. The criterion measure for assessingthe overall effectiveness was the amount of"seconds due to sewing" a garment (forexample, time on overall task) for eachsubasseribly group in the entire plant.

The results were the following:

The nominal grouping of operators intosubassembly teams and the implementation ofspecific goals and regular feedback enhancedthe social and task-related interactions andperformance of operators.

Increasing feedback and goal specificityserved as a clear focus for groupperformance improvements.

Though feedback and goal specificity may notimprove work attitudes, use of goals andfeedback substantially improved performance.Feedback was useful for initiating groupconcepts because it enabled operators topredict what was expected of them, stimulatinga greater emphasis on goal achievement.

6

Erez, M., & Arad, R. (1986). Participative goal setting:Social, motivational, and cognitive factors. Journal ofApplied Psychology, 71(4), 591-597.

Purpose

Participants

Method

Results

To study the social, motivational, andcognitive factors of participation onperformance.

96 white collar employees.

Two levels (high, low) and three types ofinvolvement ( goal setting, group discussion,and information sharing) were used.Participants evaluated job application formsfor specific job requirements. Aquestionnaire measured perceived socialinteraction, perceived involvement in goalsatting, and perceived amount of relevantinformation.

Overall performance was higher in the high-group discussion condition than the low-group discussion condition.

Number of errors corrected was higher in thehigh-group discussion and high-groupinvolvement conditions than in the low-groupdiscussion and low-group involvementconditions.

Participants performed lower in thecombination of low-group discussion, low-group involvement, and low-groupinformation, than the high-group conditions.

The percent recalled was higher in the high-group discussion conditions than low-groupdiscussion conditions.

Goal acceptance was higher in the high-groupdiscussion condition than the low-groupdiscussion condition.

Discussion Performance, incidental learning, goalacceptance, group commitment and satisfact:..onwere higher for participants in groupdiscussions. Involvement in goal setting hada significant effect on performance, thoughthe level of goal difficulty was constant.

V)8

Sorenson, J. (1971). Task demands, group interaction, andgroup performance. Sociometry, 34(4), 483-495.

Purpose

Participants

Method

Results

Discussion

To study task demands, group task behavior,and group performance.

28 three-man groups of college males.

Two types of intellectual tasks, productionand problem-solving tasks, were given to thegroups. Five task behaviors were recorded byevaluators. Group products were rated on twodimensions: quality and originality.

The results were the following:

Production tasks led groups to morestructuring, generating, and requesting thanproblem-solving tasks.

Production tasks initiated more requests forinformation and clarification, due to thehigher need to structure tasks.

n Problem-solving tasks produced biglaerquality solutions than production tasks inthe areas of task demands and groupperformance.

This study showed the relationship betweentask behaviors and performance levels on bothproduction and problem-solving tasks. Greatertaik demands and quality were related tobehaviors on a production task. Solutionoriginality was not related to behaviors on aproblem-solving task. While the group may beinvolved in problem solving, task demands mayconstrain the quality of the solution.

9

Kabanoff, 134., & O'Brien, G. (1979). The effects of task typeand cooperation upon group products and performance.Organizational Behavior and Human Performance, 23, 163-181.

Purpose

Participants

Method

Results

Discussion

To study the direct and interactive effects oftask type and structure on group products.

24 three-person groups.

Each group performed three tasks whileemploying one of four work processes. Eachtask required two levels of coordination (ideacontribution and systematic organization ofideas)/ two levels of collaboration (ideacontribution and an unsystematic organizationof ideas), and three types of tasks(production, discussion, and problem solving).

Each group responded to three criticalincidents describing medical workers in anative clinic. After each incident, fourpossible explanations for the events describedwere given. Groups selected the mostappropriate explanation for each incident anddiscussed their reasons.

The results were the following:

Collaboration yesulted in products withreduced length, originality, issueinvolvement, adequacy, quality, andcreativity.

Coordination resulted in products that werelonger, more issue involved, adequate,creative, and higher quality.

Production tasks were performed better ID),collaborative groups; problem solving taskswere performed at an intermediate level.

The results showed that when information isshared, collaboration among group membersenhanced performance. Coordinated structuresmore creative products. Collaborativestructures resulted in less adequate products.

10 12

Campbell, D., 4 Gingrich, K. (1986). The interactive effectsof task complexity and participation on taskperformance: A field experiment. Oraanizational Behaviorand Human Decision Processes, 38, 162-180.

Purpose

Participants

Method

Results

Discussion

To investigate group participation onperformance of a complex task.

40 entry-level programmers placed in groups.

Grelup3 were asked to design and write anactual computer program. Participantsanalyzed the problem, designed the program,composed the program code, and tested theprogram to meet design specifications.

The programs were distributed among foursupervisors, with each supervisor having a mixof both participation and assignment programs.Two types were identified: simple programswhich required 40 hours or less and complexprograms which required 40 hours or more.

The results showed that:

Complex tasks involved more discussion andexplanation than simple tasks.

* Participation under complex task conditionsled to better task performance than anyother task type and involvement method.

Group performance on complex tasks can beimproved through participation. Improvementsoccur as a result of increased quality andquantity of information. The resultsconfirmed the relationship between groupparticipation and task performance undercomplex task conditions, though not on simpletask conditions. Placing more effort into anassigned task may not be sufficient to obtaindesired performance levels.

11 3

Griffin, R. (1988). Consequences of quality circles in anindustrial setting: A longitudinal assessment. Academyof Management Journal, 31(2), 338-358.

Purpose To study the how quality circles affectindividuals and tne organization.

Participants 73 employees organized into eight qualitycircles.

Method This was a field experiment employing repeatedmeasures, with the measures at six months, 18months, and 36 months. Group membersparticipated in a two-day off-site program onproblem solving. The program was f)llowed bygroup problem solving activities. After eachtime measure, questionnaires were administeredand interviews were conducted.

Results Scores for the four primary individual-lvaldependent variables (job satisfaction,organizational commitment, performance, andintentions to quit) improved gradually up tothe 18 month mark, but subsequently decreasedto their initial baseline levels. The groupswere considered successful for the individualsand organization for about two years, but thenperformance declined to initial baselinelevels.

Discussion Though performance ratings were at theirhighest level toward the latter portion of theprogram, effectiveness of quality circlesdeclined to their original level ofperformance after program termination. Theresults suggest that quality circles must havea management system to monitor changes inquality circle performance so that exemplaryperformance is maintained, not eliminated.

Marks, M., Mirvis, P., Hacketf', SO, & Grady, J., Jr. 0986).Employee participation in a quality circle program:Impact on quality of work life, productivity, andabsenteeism. Journal of Applied Psychology, 71(1), 61-69.

Purpose

Participants

Method

Results

Discussion

To study employee participation, quality cfwork life, productivity, and abseriteeism in aquality circle program.

109 workers .n a manufacturing company.

Fifty-three of the 109 eligitale emplayeeselected to participate in the program and weretrained and placed into one of six qualitycircles. The groups 11,..?t one hour per week tosolve work-related problems. Nominal groups,brainstorming, cause-efect analysis, andfishbone problem analysis flow charts wereused groups. A questionnaire was administeredto participants and nonparticipants after twoyears of program implementation.

The results showed that productivity improvedfor participants during the 24 month periodfollowing implementation of the program.Improvements were noted in: percentage ofhours spent on production, plant efficiencyrates, and hourly productivity rates.

The results indicated how participatior inquality circles affects attitudes and employeeproductivity. The results support theconclusion that participants perform produceat a higher rate than non-programparticipants.

1315

Buller, POI & Bell, C., Jr. (1986). Effects of team buildingand goal setting on productivity: A field experiment.Academy of Management Journal, 29(2), 305-328.

Purpose To study the effects of team building on thedevelopment of group strategies.

Participants

Method

Results

Discussion

53 miners in an underground metal mine.

36 miners were assigned to undergo teambuilding in order to work in a stope (a groupof miners who work together in a smallunderground area): 12 stopes were assigned forthe team building condition, while the eightremaining stopes were assigned to the non-teambuilding condition. Productivity was the unitof analysis.

Stope miners and their shift bosses setspecific, difficult, and attainable goals fortons per manshift and grades of ore collectedfor three months. Miners and bossesestablished mutually agreed upon goals thatmet the stated criteria. Shift bossesprovided weekly feedback sheets to miners inthe goal-setting condition.

The results showed that team building had apositive effect on the grades of silver andlead collected. Goal setting improved tonsper manshift collected. Structured interviewsshowed that the miners preferred the team-building and goal-setting interventions.Interviews showed that most miners expressedideas to management openly. Miners mentionedthat the feedback helped them increase thequality of ore collected.

The findings showed that goal setting andfeedback can serve as important ways toinfluence group performance. Miners whoreceived feedback on their goals did notnecessarily work harder than miners who didnot receive feedback on their goals. However,-.hey did report using more appropriate ways toattain the task.

Latham, G.P.r & Marshall, H.A. (1982). The effects of self-set, participatively set, and assigned goals OA theperformance of government employees. PersonnelPsychology, 35, 399-404.

Purpose

Participants

Method

Results

Discussion

To study assigned versus participative goalsetting.

57 supervisors in a government agency.

All participants were told that a job analysiswas being conducted to define effectivesupervisory behavior. Each participantbrainstormed job behaviors that made thedifference between effective and ineffectivejob behavior as a supervisor. Goals were setfor the number of behaviors written within a20 minute period.

There was no difference in goal difficultybetween participative goal-setting andindividual goal-setting. Goal difficulty washeld constant between the participative andassigned goal conditions by imposing a goalagreed upon by an employee in theparticipative condition upon an employee inthe assigned condition. There was nodifference among the three goal-settingconditions in goal acceptance or performance.

Previous research has shown that specific anddifficult goals lead to higher performancelevels. These results support the notion thatthe process used to set goals may be asimportant as the goals themselves.

15 17

Bottger, P.C., & Yetton P.W. (1987). Improving groupperformance by trai. ing in individual problem solving.Journal of Applied Psychology, 72(4), 651-657.

Purpose To study how training improves individualgroup members' use of task knowledge.

Participants 169 managers and 207 MBA students working in80 groups.

Method The task was the "Moon Survival" problem,which required participants to imaginethemselves crash landed on the moon 200 milesfrom the home base. Participant interviewsconfirmed that the exercise was perceived astesting task knowledge and small groupdiscussion skills.

Results Group performance was perceived to be betterthan individual performance on problemsolving. The results also showed thatindividual problem solving complimented thegroup problem solving. The results supportthe notion that group performance is stronglyinfluenced by individuals' task ability.

Discussion Individual performance improved groupperformance on problem-solving. In addition,by training the participants in effectivesearch and evaluation routines prior toassembling in groups, individuals use of taskknowledge was upgraded and team achievementwas enhanced.

16

Tang, T.14., Tollison, P.G.1 & Whiteside, H.D. (1987). Theeffect of quality circle initiation on motivation toattend quality circle meetings on task performance.Personnel Psychology, 40, 799-813.

Purpose To study motivation to attend quality circlemeetings and task performance.

Participants

Method

Results

Discussion

47 quality circles in an assembly plant.

Four variables were examined, quality circle:attendance rate, ability to solve problems,and member performance.

The results were the following:

peer quality-circle initiation resulted inhigher attendance at meetings thanattendance at management quality-circleinitiations.

higher attendance rate was shown in meetingsassociated with a low quantity of projects,but higher project completion rates.

How organizations initiate quality circlesinfluences the future behavior of groupmembers. Peer quality-circle initiationresulted in higher attendance. Managementquality-circle initiation resulted in lowerattendance. How to influence employees toinitiate quality circles remains an issue inmany organizations.

Hall, J., & Williams, M.S. (1970). Group dynamics trainingand improved decision making. Journal of AppliedBehavioral Science, 6(1), 39-68.

Purpose

Participants

Method

Results

Discussion

To study how training on the decision-makingprocess affects group performance.

377 individuals, formed into 60 groups.

30 groups trained in group dynamics werecompared with 30 untrained groups in theirperformance on the "Twelve Angry Men" decisionmaking task. Three different populations ofdecision makers were studied: collegestudents, managers, and psychiatric workers.

Trained groups consistently performed moreeffectively than untrained groups on measuresof decision quality, use of resources, andcreativity.

The importance of this study is somewhatlimited by the special nature of the task andby the possible extraneous factor of usingparticipants from different populations. Theresults demonstrated that group performancewas improved under the trained-conditionsversus the untrained condition.

1820

Pritchard, R.D./ Jones, S.D., Roth, P.L., Stuebing, K.K, &Ekeberg, S.E. (1988). Effects of group feedback, goalsetting, and incentives on organizational productivity.journal of Applied Psychology, 73(2), 337-358.

Purpose

Participants

Method

Results

To measure productivity for complex jobs andanalyze differences between group-levelfeedback, goal setting, and incentives.

85 service personnel at an Air Force base.

The method consisted of a baseline period ofeight to nine that was followed by a fivemonth period of feedback based on theProductivity Measurement and EnhancementSystem (ProMES). Goal setting was added tofeedback for five months. Incentives, whichincluded time off from work, were also addedto feedback and goal setting after anotherfive months.

The results were the following:

w Group-level feedback increased productivityan average of 50 percent over baseline.

a Group-goal setting increased productivity75 percent over baseline.

11 Group incentives increased productivity 76percent over baseline.

Control group data showed only a slightincrease in i-rcductivity over the sameperiod, and the performance Level ofpersonnel either stayed the same ordecreased.

Job satisfactiJn, turnover intentions, andmorale, were as good or better followingthe interventions.

19

Discussion The results demonstrated the importance offeedback for gtoups in organizations wherethere are rapid technological or marketchanges. However, it is unclear how muchparticipation is actually netessary fordeveloping effective group feedback systems.

Libby/ R., Trotman, K.T./ Zimmer/ I. (1987). Membervariation/ recognition of expertise/ and groupperformance. journal of Applied Psychology, 72(1)/ 81-87.

Purpose

Participant

Method

Results

To study the variation in group performance torecognize expertise among group members.

60 loan officers in banks and financecompanies.

A task was distributed to individuals whocompleted the task independently. Individualsthen were assigned to three-person, prior toreceiving feedback on their individualperformance. Participants then completed thesame task as a group. Ability to recognizeexpertise was measured by the perceptions ofgroup members compared to actual individualperformance and judgments about who was themost influential member.

Expert loan officers serving in groupsperformed as well as the most influentialindividuals. Whether a particular groupoutperformed or underperformed its compositecould be explained by variation in groupmembers' performance and ability to recognizedifferential expertise.

Discussion The results suggest that the ability torecognize expertise and the variation inindividual performance are major determinantsof group. Ability to recognize expertise andvariations in individm.1 expertise areaffected by the type and complexity of thetask. The results also suggest that the typeof task will affect performance.

21 23

Hollenbeck, J.R.1 & Brief, A.P. (1987). The effects ofindividua2 differences and goal origin on goal settingand performance. Organizational Behavior and HumanDecision Procesces 40, 392-414.

Purpose To study how individual differences and goalorigin affects goal setting and performance.

Participants

Method

Results

Discussion

102 college students.

Participants solved a set of anagrams ingroups of six to eight. They were given a oneminute practice trial to unscramble as many aspossible. The total time for the experimentwas ten minutes.

The results were as follows:

s In self-set goal.condition, self perceptionswere related to the difficulty of the goalsselected, with more difficult goals set byindividuals having higher perceptions oftheir ability.

In assigned-goal conditions, there was arelationship between selection of goaldifficulty and performance on the task, forparticipants having high self-esteem.

The findings suggest that for groups high inself-esteem, more difficult goals serve toincrease perceptions of task-specific ability,which in turn enhances performance.

22 24

Hollenbeck, J.R., & Williams, C.R. (1987). Goal importance,self-focus, and the goal setting process. Journal ofApplied Psychology, 72(2), 204-211.

Purpose To study the effects of goal importance andself-focus on goal-setting.

Participants

Method

Results

143 salespersons in a department store.

The study used three data collection methodsto examine the importance of goals forsalespersons: archival records, questionnaireresponses, and a policy-capturing exercise.

The results showed that perceived pastperformance on goals was much stronger forhigh self-focused salespersons than for lowself-focused salespersons.

Discussion: The results confirm that perceptions of pastperformance influence future performance and,thus, might be a good predictor of actualperformance.

2325

Garland, H. (1983). Influence of ability, assigned goalz:, andnormative information on personal goals andperformance: A challenge to the goal attainabilityassumption. Journal of Applied Psychology, 68(1), 20-30.

Purpose

Participants

Method

Results

Discussion

To show how information about assigned goaldifficulty affects the goals individuals setfor their own performance.

58 undergraduate students.

Participants weee assigned to one of sixconditions representing two levels of goaldifficulty (easy versus difficult) and threelevels of normative information. Participantsset goals on a creativity task, which wasperformed over ten repeated trials.Participants then selected whether they wantedto continue working on more trials.

The results were the following:

Personal goals were influenced by assignedgoals, but ability had not related topersonal goals.

Information did not influence personal goalsdirectly, but did interact with assignedgoals.

.1 Assigned goals and ability affectedperformance.

Assigned goals influenced performance,indirectly through their influence onpersonal goals and directly through personalgoals assigned at a difficult level.

The results were consistent with other studiesthat showed specific and difficult goalsmotivate individulls and groups to achievehigher performance than do easier goals.

24

A. Ahad Osman-Gani is a doctoral research associate inthe graduate program in Training and Development, The OhioState University.

Shashi B. Gowda is a staff member of an internal humanresource development consulting department, XeroxCorporation, Leesburg, Virginia.

Patrick E. Smylie is a doctoral research assistant intIle graduate program in Training and Development, The Ohio0L:ate University.

Ronald L. Jacobs is associate l_rofessor and coordinatorof the graduate proaram in Training and Development, The OhioState University, 160 Ramseyer Hall, Columbus, Ohio 43210,(614) 292-5037.

27