key words Flashcards
1
Q
chi-square test
A
- test of difference with unrelated data
- nominal level data
- indipendent measures design experiment
2
Q
mann-whitney U test
A
- test of difference with unrelated data
- ordinal level data
- independent measures design experiment
3
Q
binomial sign test
A
- test of difference with related data
- nominal level data
- repeated measures design experiment
- matched participant design experiment
4
Q
wilcoxon signed ranks
A
- test of difference with related data
- ordinal level data
- repeated measures design experiment
- matched participant design experiment
5
Q
spearmans rhoe
A
- test of correlation
- ordinal level data
6
Q
independent t-test
A
- interval or ratio level data
- test of difference with unrelated data
- independent measures design experiment
7
Q
related t-test
A
- test of difference with related data
-interval or ratio level data - repeated measures design experiment
- matched participant design experiment
8
Q
pearson’s product movement
A
- test of correlation
-interval or ratio level data
9
Q
type 1 error
A
- ‘false positive’
- the alternative hypothesis is mistakenly accepted
10
Q
type 2 error
A
- ‘false negative’ errors
- the alternative hypothesis is mistakenly rejected
11
Q
=
A
equal to
12
Q
<
A
less than
13
Q
«
A
much less than
14
Q
>
A
greater than
15
Q
> >
A
much greater than
16
Q
∝
A
- is proportional to
Proportionality (i.e. a correlation has a linear property, such that as one co-variable increases by a certain amount [e.g. it doubles] so the other co-variable will increase by the same amount)
17
Q
~
A
Approximately
18
Q
reliability
A
to the consistency of a test or measure.
19
Q
internal reliability
A
- The consistency of a measuring device
- the procedure is standardised and consistent which can be replicable with all participants
20
Q
external reliability
A
- there is enough participants to establish a consistent effect
- the same findings would be obtained if the study was repeated
21
Q
ways to check the reliability of a test or study
A
- split-half method
- test-retest method
- inter-rater reliability
22
Q
split-half method
A
- compare the score from one half of the questions to the score for the other half of the questions to see if the participants scored consistently on both halves
23
Q
test-retest method
A
- giving the participants the same test/measure at a different point in time to check whether their two scores are consistent
24
Q
inter-rater reliability
A
- Two or more observers record the behaviour, and then their results are compared to check the level of agreement in their results
- (a high correlation between their scores of 0.8 or more would indicate high inter-rater reliability)
25
internal validity def
whether the test itself is accurately measuring what it
intends to
26
internal validity
- face validity
- concurrent validity
- criterion validity
- construct validity
27
face validity
Whether a test appears (on the face of it) to be measuring what it intends to
28
concurrent validity
Where a test or study measure gives the same results as another test or study that is measuring the same concept
29
criterion validity
Refers to how much one test or measure predicts future performance on another test or measure
30
construct validity
Refers to whether a test or study actually measures the concept it sets out to measure (and extraneous variables are controlled for)
31
external validity
refers to whether the research can be generalised to
different people or situations
32
external validity
- population validity
- ecological validity
33
population validity
Refers to the degree to which the sample used in the research is representative of a diverse group of people (of different ages, genders, occupations, education levels, etc.)
34
ecological validity
Refers to how accurately a piece of research reflects real-life situations
35
representativeness
- refers to the sample used in the research
- if the sample is diverse and includes people from different ages, genders, occupations, education levels, etc., it will be more representative of the (target) population
36
generalisability
- refers to the results of the research
- if the sample used in the research is biased and not very diverse, the results cannot be generalised to everyone in the target population.
37
demand characteristics
- when participants work out the aim of the research either because it is obvious, or as a result of a repeated measures design being used.
- They may then change their behaviour and act in the way they think the researcher wants them to act.
38
social desirability
- when participants change their behaviour to present an image of being a good member of society or to fit into social norms, rather than showing their true behaviour.
39
researcher/observer bias
- refers to the way the researcher collects and interprets the results of research.
- They may interpret behaviour based on their prior expectations and therefore this would lower the validity of the findings.
40
researcher/observer effects
- refers to the way that participants’ behaviour is influenced by the presence (and their characteristics) of the researcher
41
ethical guidelines (broad guideline)
- **R**espect
- **I**ntegrity
- **C**ompetence
- **R**esponsibility
| RICE (E=R)
42
respect
- informed consent
- right to withdraw
- confidentiality
43
competence
researchers need to operate within their capabilities and ot give any advice beyond that which they are qualified to give
44
responsibility
- protection from harm
- debrief
45
integrity
- deception
46
format to follow when publishing an article in academic journals
1. **Name of Author** (the surname is followed by the initials of the first names
2. **(Date of publication)** of publication of the article (in brackets)
3. **'article** **title'** (in single inverted commas
4. ***Journal title*** (in italics)
5. **volume** of the journal
6. **issue number** of the journal
7. **page range** of article
| **N**EVER **D**ATE **T**EACHERS **J**UST **VIP**S
47
example of the format used when publishing an article
Moray, N. (1959) ‘Attention in dichotic listening: Affective cues and the influence of instructions’, *Quarterly Journal of Experimental Psychology*, 11 (1), 56-60.
48
peer review
- academic articiles need to be read and evaluated by experts in the same field before being published so that they can ensure that the methodology used is robust
- (ie. valid and reliable measures have been used to collect the results)
49
strengths of peer review
- can be used to check that research will be useful before it is funded
- ensures only the most relevant and robust research is published
- ensures that only valid results are published so the journals retain their reputation
50
weaknesses of peer review
- can take a long time
- some reviewers may not pass research that contradict their own
- may not be possible to detect research that has used false data
51
the study of cause-and-effect
- where a researcher can show that one variable is actually causing a change in another variable
52
falsifiablity
The ability, in principle, to prove a claim wrong
53
Replicability
The ability to repeat a study and therefore test to see if its findings are reliable (the use of controls and standardised procedures make it more replicable)
54
Objectivity
When a claim is a matter of fact, rather than opinion
55
Induction
Empirical research is carried out and then a theory is developed to make sense of findings
56
Deduction
A theory is developed and then empirical research is carried out to see if the theory is correct (i.e. supported by evidence)
57
Hypothesis testing
Based on a psychological theory, a prediction is made about how participants would be expected to behave, which can be tested through research (e.g. experiment, observation, etc.)
58
Manipulation of variables
When an independent variable is changed (manipulated) to see what effect this has on a dependent variable (how it affects behaviour).
59
Control
This is imposed on experiments to ensure that results are due to the independent variable, rather than extraneous variables
60
Standardisation
The test conditions are kept the same for all participants
61
Quantifiable measurements
The use of numerical data, which can be used to compare between conditions. This should be observable and objective.
62
Interval
or ratio
- highest level of data
- Analysis is made of the scores achieved by individual participants.
- involves the use of standard universal scales (e.g. seconds, kilograms, metres, etc.). The sizes of the gaps between (say) the highest score, second highest score, third highest score, etc., are taken account of.
- e.g. crossword, participants are recorded in minutes and seconds for how long it takes them to complete the crossword
63
ordinal
- medium level of data.
- Analysis is made of individual scores achieved by participants, but only in relation to each other
- what is analysed is their rank position within a group, rather than their actual scores
- . No account is taken of how much further highest is from second highest, etc., so the measures used may not be carefully calibrated.
- e.g. crossword, at the end of 15 minutes a note was made of how many correct answers each participant has entered to the crossword in that time, these are placed in rank order.
64
nominal
- lowest level of data
- It is a ‘headcount’ of the number of participants who do one thing as opposed to another.
- e.g. Headcount of the number of males successfully completing a crossword in 15 minutes as opposed to the number of females successfully completing a crossword in 15 minutes.
65
level of data collected in piliavin
**nominal**: resarcher recorded the number of times each victim was helped which would have been recorded as a tally
**interval/ratio**: researchers also recorded the median time taken to help for the drunk and ill victim in seconds
66
level of data collected in ball and bucket
- ordinal: recorded the number of balls each participant was able to successfully throw into a bucket. their scores ranked in order from highest to lowest
67
level of data collected in milgram
- nominal: recorded the number of partcipants who were obedient and the number of participants who were disobedient
68
level of data collected in loftus and palmer
- **interval/ratio:** in experiment 1, participanst estimated the speed of the vehicles in miles per hour
-** nominal:** in experiment 2, they recorded the number of participants who said they did/did not see broken glass in the video clip
69
level of data collected in lee
- ordinal: after each story, the children ratd the behaviour of the character on a 7 point scale (from very very good to very very naughty)
70
level of data collected in chaney
- nominal: in the questionnaire, participants were asked yes/no questions about whether they had used the inhaler/funhaler the previous day and whether they screamed when using the mask, whether they reacted with pleasure while using it etc.
71
strengths of nominal data
- Better than ordinal and interval because:
- Quick and easy to obtain because it is just a headcount
- Can be displayed in pie charts (which can be easily made sense of)
72
strengths of ordinal data
- Better than nominal because:
- Can calculate mean, median and mode as measures of central tendency (so more detailed)
- Can also calculate measures of dispersion
- Can calculate individual scores of participants and see how they differ
73
strengths of interval/ratio
Better than nominal because:
- Can calculate mean, median and mode as measures of central tendency
- Can also calculate measures of dispersion
- Can calculate individual scores of participants and see how they differ
Better than ordinal because:
- Scores can be compared directly as precise values are recorded (i.e. you can see the actual difference between scores rather than just the rank position)
- The scores are more consistent as the same universal scale is used (e.g. a cm is always measured in the same way)
74
weaknesses of nominal data
- Worse than nominal and interval because:
- Can only analyse the mode of data and cannot calculate the mean or median
- Cannot analyse measures of dispersion (such as range and standard deviation)
- Less precise as data is grouped into categories (we don’t know how individual participants scored)
75
weakness of ordinal data
- Worse than interval because:
- Ordinal data can be subjective (as people may interpret rating scales differently)
- Although we can work out the rank order of participants, we don’t always know the exact difference between individual scores
- Worse than nominal because:
- More time consuming and complex to analyse
76
weakness of interval/ratio data
- Worse than ordinal because:
- Can only be used with concepts that are measurable through universal scales (can’t be used with attitudes, opinions, etc.)
- Worse than nominal because:
- More time consuming and complex to analyse
