Research Methods Flashcards
(101 cards)
1
Q
Types of validity
A
Face Construct Concurrent Criterion Population Ecological
2
Q
Binomial sign test
A
Repeated measures
Nominal data
3
Q
Chi-squared
A
Independent measures
Nominal data
4
Q
Mann-Whitney U
A
Independent measures
Ordinal data
5
Q
Alternate hypothesis
A
There WILL be a significant difference
6
Q
Null hypothesis
A
There WON’T be a significant difference
7
Q
One-tailed hypotheses
A
Directional
8
Q
Two-tailed hypotheses
A
Non-directional
9
Q
Spearman’s Rho
A
Correlation
10
Q
Unstructured interview
A
No pre-determined questions
✅ detailed, more comfortable
❌ time consuming, may go off topic, open to bias
11
Q
Closed questions
A
Restricted response
✅ easy to analyse
❌lacks detail
12
Q
Negative correlation
A
As one increases the other decreases
13
Q
Response bias
A
When people chose the middle option
14
Q
Criterion validity
A
Refers to the extent the test scores can predict a future behaviour
15
Q
Construct validity
A
Whether a measure is assessing it’s underlying constructs
16
Q
Population validity
A
Evaluated whether the sample represents the entire population
17
Q
Ecological validity
A
Evaluated whether the results of a study can be generalised to real life situations
18
Q
Naturalistic observation
A
Carried out in a natural environment where you’d expect to see that behaviour occur
✅ ecological validity
❌ less control
19
Q
Structured observation
A
The researcher has a behaviour checklist to tally
✅ clear direction
❌ behaviours that aren’t on the checklist don’t get recorded
20
Q
Participant observation
A
Where the researcher is apart of the group that is being observed
✅ gain first hand data
❌ observer bias
21
Q
Non-participant observation
A
Researcher not apart of the group being observed
✅ less chance of observer bias
❌ miss certain behaviours
22
Q
Overt observation
A
P’s know they’re being observed
✅ ethical
❌ demand characteristics
23
Q
Covert observation
A
P’s don’t know they’re being observed
✅ no demand characteristics
❌ less ethical
24
Q
Observer bias
A
Where they interpret the behaviour they want
❌ less likely to measure what intends to
25
Random sampling
Equal chance of being chosen
✅ unbiased
❌ time consuming, less willing
26
Order of a report
```
Abstract
Introduction
Method
Results
Discussion
Bibliography/references
Appendices
```
27
Wilcoxon’s test
Repeated measures
| Ordinal data
28
Report: abstract
```
General aims/hypothesis
Sample
Design/procedure
Results
Conclusions
```
29
Repeated measures
P’s complete all conditions
✅ don’t have to find double the amount of p’s, no individual differences
❌ experience order effects, demand characteristics
30
Matched pairs
P’s from different conditions are matched
✅ avoids order effects, less chance of demand characteristics
❌individual differences, more p’s
31
Lab
Controlled conditions, manipulate IV
✅ easily replicated
❌ lacks ecological validity
32
Snowball sampling
Asking p’s to nominate another person
✅ find people with rare characteristics
❌ time consuming, bias
33
Opportunity sampling
Selecting people who are readily available at the time
✅ quick, easy, cheap
❌ biased, unrepresentative
34
Volunteer sampling
Select themselves to participate
✅ p’s willing
❌ time consuming, unrepresentative
35
Ethnocentrism
When a sample is taken from one culture/area
36
Nominal data
Where data is in separate categories
37
Ordinal data
Data is ordered in some way
38
Measures of dispersion
How spread out the data is
39
Standard deviation
Measures how spread out the data is from the mean
High - very spread out from the mean
Low - close to the mean
0 - all the data values are the same
40
% of 1 standard deviations
% of 2 standard deviations
% of 3 standard deviations
1 - 68%
2 - 95%
3 - 99.7%
41
Field
In the field observing, manipulate IV on DV
42
Quasi
IV not manipulated, naturally occurring
✅ no demand characteristics
❌ difficult to see cause and effect
43
Structured interview
Pre-determined questions
✅ clear direction, easy to replicate
❌ may not ask questions as it’s focused on set questions, feel too formal
44
Semi-structured interview
Pre-determined questions but can go off topic if necessary
✅ flexible
❌ can go off topic, time consuming
45
Open questions
No restricted response
✅ gives reasons, detail
❌ difficult to analyse
46
Likert scale
Indicate how much they agree/disagree with a statement
✅ easy to analyse
❌ response bias
47
Correlation
How strongly 2 variables are related to each other
✅ strength in relationship, may suggest further research
❌ cannot assume cause and effect, don’t always get details behind data
48
Independent measures
P’s complete 1 condition
✅ no order effects, no demand characteristics
❌ individual differences, need more p’s
49
Positive correlation
Both co-variables increase/decrease together
50
No correlation
They are uncorrelated
| No relationship
51
Correlation coefficient
Refers to a number between +1 and -1 and states how strong a correlation is
+1 positive correlation
-1 negative correlation
Near 0 weak correlation
0 no correlation
52
Social desirability bias
People change answer as they want to look good
53
Threats to validity
```
Observer bias
Researcher bias
Demand characteristics
Social desirability bias
Order effects
Individual differences
Response bias
```
54
Face validity
Whether the study is measuring what it appears to on the surface if it
55
Concurrent validity
Comparing the results of one measurement to another
56
Unstructured observation
The researcher doesn’t have a pre-determined behaviour checklist
✅ not restricted
❌ lose sight of actual observation
57
Controlled observation
In a controlled setting, researcher controls who they observe and p’s often know
✅ more control
❌ demand characteristics
58
Behavioural categories
Where there are categories of behaviour/checklist to observe
✅ gives researcher direction
❌ restricted
59
Coding frames
Where behavioural categories are coded and can be rated for severity
✅ contain more detail
❌ observer bias
60
Inter-rater reliability
2 or more observers observe the same behaviour, time, behaviour checklist and compare data
✅ can see if they recorded data in similar way
❌ time consuming
61
Event sampling
Researcher recording a behaviour every time it happens
✅ less likely to miss behaviours
❌ difficult to keep track
62
Time sampling
The observer records what the participant is doing in fixed time categories
✅ less likely to miss behaviours
❌ miss some behaviours
63
Observer effects
When the presence of an observer in an overt observation changes the behaviour of the p’s
❌ demand characteristics
64
Report: introduction
Background research
| 2 hypothesis
65
Report: method
Design
Participants
Materials
Procedure
66
Target population
Group you’re trying to find out about
67
Interval data
Where units of equal measurements are used
68
Variance
Measures how much a set of values is spread out
69
Report: results
Descriptive statistics
| Inferential statistics
70
Report: discussion
Draw conclusions from results
Evaluate experiment
Consider modifications for future research
71
Report: bibliography/ references
Reference from any source used
| The background research from introduction
72
Report: appendices
Relate to method and results
Materials
Consent form
Debrief
73
Questionnaires
✅ large numbers of participants, quick
❌ low response rates, less flexible, bias
74
Rating scales
Represent thoughts/feelings using a number
Not at all 1 2 3 4 5 a lot
✅ easy to analyse
❌ response bias
75
Semantic differential scale
Measures a person’s attitudes towards something by putting something on a scale between 2 descriptive words
✅ easy to analyse
❌ response bias
76
Harvard referencing
Milgram.S (1963) Behavioural Study of Obedience. Journal of Abnormal and Social Psychology (in italics), 67, 371-378
Name, date, title of paper, journal published in, volume number, pages
77
Levels of measurement
Nominal
Ordinal
Internal
78
Raw data
Raw data = primary data (collected first hand)
79
Descriptive statistics
```
Mean
Median
Mode
Variance
Standard deviation
Range
```
80
Measures of central tendency
Mean
Median
Mode
81
Measures of dispersion
Variance
Standard deviation
Range
82
Graphs
```
Bar chart - categories
Scattergraph - correlation
Pie chart - categories
Line graph - compare 2 conditions
Histogram- continuous data
```
83
Range
Biggest - lowest + 1
84
Standard deviation calculation
```
Calculate mean
Subtract mean from each value
Square each result
Calculate squared differences (no. Of values -1)
Square root
```
85
DAJO: own experiences
Observation - student behaviour in cafe, observing shopping behaviour in a mall
Experiment - picture vs words
Correlation - height vs shoe size
Self-report - anything
86
Degrees of freedom
(Number of rows of data - 1) x (number of columns of data - 1)
87
Quantitative data
Data in form of numbers
✅ easy to analyse, make comparisons
❌ doesn’t give context, can distort truth
88
Qualitative data
Data in the form of words
✅ depth, less open to bias
❌ hard to analyse, open to interpretation
89
Type 1 error
False positive
Wrongly accepted alternate instead of accepting the null
90
Type 2 error
False negative
Wrongly accepted null instead of accepting the alternate
91
Probability
The likelihood of an event of numbers occurs due to chance
P<0.05
5% chance
95% significant
Accept alternate
92
Primary data
Researcher collects themselves
✅ fits needs of experiment
❌ time consuming, costs more
93
Secondary data
Collected by somebody else
✅ saves time, saves money
❌ not be appropriate for needs of study
94
Critical value
A critical value is a figure in a table of critical values that the answer from an inferential statistical test is compared with to check if the findings are statistically significant
95
Ranking
Test scores would be assigned a number, using rank 1 for the lowest score, rank 2 for the next score and so on for both the maths test and the physics test (but each set of scores ranked separately to each other).
96
Internal validity
Is the study measuring what it intends to within the study itself?
97
External validity
The extent that the results from a study that can be generalised beyond the current study
98
Independent variable
This is the variable/thing that the researcher wants to change (manipulate) to see what effect is has on something else.
99
Dependent variable
The variable the researcher measures.
100
Extraneous variables
Any variables controlled by the researcher.
101
Confounding variable
Any variables NOT controlled by the researcher that could ruin the results
