Date post: | 28-Dec-2015 |
Category: |
Documents |
Upload: | megan-tucker |
View: | 214 times |
Download: | 1 times |
Overview
• What is a scientific inquiry?• What are the types of studies in
psychology?• How is data collected?• How is data analyzed?
What is a scientific inquiry?
4 basic principles– Description
• What is it?
– Prediction• When does it happen?
– Causality• What causes it no happen?
– Explanation• Why does it happen?
Overview: stages in research
Formulate Hypothesis
Design Study
Collect Data
Analyze Data
Disseminate Results
Major Study Designs
• Experimental Design– Manipulate conditions
• Correlational Design– Study the relationships between
variables
• Descriptive Study– Observe and describe a phenomenon
• Manipulate independent variables• Measure dependent variables• Rule out confounds and alternative
interpretations
• Establish causal relationships
The Experiment
Operational definitions
• Explains how a variable will be manipulated and measured
• Example: an operational definition for the hypothesis ‘television leads to increased aggression’ is – the variable ‘television’ will be studied by exposing children to a horror film or a family film; aggression will be measured as the number of times a child hits another child during the 60-minute play period following the film. Alternatively, in a survey we can ask parents how many hours/week their kids watch TV and ask them to rate their aggressiveness. This is a different set of definitions
Hypothesis: Gum is bad for your teeth
• Gum: – the number consumed weekly by a
person
• Bad for teeth:– Option I: number of annual visits to the
dentists– Option II: number of dents/fillings
Hypothesis: Males are angry more often than females
• Gender– Ask respondents to indicate whether
they are males or females
• Angry– Expose them to an annoying situation
and measure facial expressions of anger
Hypothesis: Noise is bad for your hearing
• Noise– Average level of noise in work place– Number of hours/week spent in
environments louder than 90 dB-SPL
• Hearing– Thresholds for frequencies 250-8000 Hz
measured with an audiometer
Major components of an experiment
• Independent Variables (IV’s)• Dependent Variables (DV’s)• Random Sampling from
Representative Populations• Random Assignment to Conditions
or Treatments• Efforts to “Control” or minimize
extraneous/irrelevant factors
The correlational design
• Correlational studies tell us how variables are related or associated
• A correlation reflects the degree of association between two variables
Correlations are expressed with scatter plots
0 10 20 300
20
40
60
0 10 20 300
20
40
60
0 10 20 300
0.5
1
1.5
2
0 10 20 300
20
40
60
X
Y
r = 1 r = -1
r= 0 r = 0.85
Correlations in ‘real life’
0
20
40
60
80
100
0 2 4 6 8 10
0
20
40
60
80
100
0 2 4 6 8 10
Working memory (score)
Fre
quen
cy d
iscr
imin
atio
n w
ith s
tand
ard
(JN
D,
%)
Fre
quen
cy d
iscr
imin
atio
nw
ithou
t st
anda
rd (
JND
, %
)
r = 0.73 r = -0.17
Problems with inferences based on correlations
• The third variable problem• The directionality problem
Example
• Directionality:– Does poor working
memory cause poor frequency discrimination of vice versa?
• Third Variable:– Both could be
affected by another factor (e.g. I.Q.)
0
20
40
60
80
100
0 2 4 6 8 10
Working memory (score)
Fre
quen
cy d
iscr
imin
atio
n w
ith s
tand
ard
(JN
D,
%)
r = 0.73
Correlation ≠ Causation
יוני רוצה להיות יותר גבוה. כל יום אחרי •הלימודים הוא מתמתח ומתמתח בתקווה
חודשים הוא אכן גבוה יותר...6שיגבה. אחרי
האם המתאם בין התמתחות וגובה הוא חיובי •או שלילי?
מהי ההשערה המיחסת סיבתיות למתאם?•
אילו גורמים אחרים יכולים היו לגרום ליוני •לגבוה?
Correlation ≠ Causation
חוקר מעביר לכל תושבי חיפה מבחן •במתמטיקה. הוא מגלה שלאנשים גבוהים
יותר, יש ציונים גבוהים יותר.
האם המתאם בין גובה וציון במבחן הוא חיובי •או שלילי?
מהי ההשערה המיחסת קשר סיבתי למתאם?•
אילו גורמים אחרים יכולים להסביר את •המתאם?
Descriptive Studies
• Observe and classify behavior (observational studies)
• Natural observation• Participant observation
Data Collection Methods
• Observing• Asking questions• Case studies• Measures of performance
– RT, JND, score…
• Measures of brain activity• Research in animals• Ethical issues are involved in each
type!
Observing in natural settings
• Unobtrusive• Often used to study
– Primate behavior– Infants– Social situations (?)
Questionnaire Studies
• Open ended• Scales of measurement
Strongly agree --------------------------------------- Strongly disagree
Positive ------------------------------------------------ Negative
Often --------------------------------------------------- Never
Measures of Performance
• Researchers can measure performance to infer underlying mechanisms– Reaction time (RT)– Response accuracy (e.g. % correct)– Detection or discrimination threshold
Example I: Stroop
red
green blue
In what color is the word printed?
red
green blue
<Congruent Incongruent
? מודפסת כל מילהצבעבאיזה
אדום ירוק אפור כחול שחור צהוב ירוק כחול אדום
אדום כחול צהוב אפור ירוק כחול שחור אדום צהוב
שחור אדום צהוב אפור ירוק שחור כחול צהוב ירוק
? מודפסת כל מילהצבעבאיזה
אדום ירוק אפור כחול שחור צהוב ירוק כחול אדום
אדום כחול צהוב אפור ירוק כחול שחור אדום צהוב
שחור אדום צהוב אפור ירוק שחור כחול צהוב ירוק
Measuring brain activity non-invasively
• Direct measure of neural activity– EEG/ERP
• Indirect measures of neural activity– PET– fMRI
Ethical issues in research
• Issues in conducting the study – informed consent etc’.
• More general issues – does the goal justify the means? Is it likely that those participating in the study are representative of those that stand to benefit from the outcomes?
Analyzing Data
• Evaluating the quality of the data– Reliability, validity and accuracy
• Summarizing the data– Descriptive statistics and correlations
• Determining if the results confirm a hypothesis– Inferential statistics
Reliability, Validity and accuracy
• Accuracy – the extent to which a measurement is free from error.
• Reliability – the extent to which a measure is stable and consistent over time and similar conditions
• Validity – the extent to which the data address the hypothesis in the way intended.