research methods Flashcards

You may prefer our related Brainscape-certified flashcards:
1
Q

when to use mann whitney u

A

independent groups
ordinal data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

what does N stand for in mann whitney u

A

N1 - sample size of group 1
N2 - sample size of group 2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

what does R1 stand for in mann whitney U

A

sum of ranks for group one/ the largest group

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

how to calculate mann whitney U

A
  • rank the scores of the entire sample together
  • put back into separate groups and add up rank total
  • larger = group 1
  • plug into equation
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

how to calculate U for the other group

A

Ub = (N1 x N2) - U

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

what makes mann whitney u significant

A

less than or equal to the critical value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

what to include in hypothesis writing

A
  • null or alternate (significant or not)
  • operationalise IV (what changes)
  • operationalise DV (how it’s measured)
  • context
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

strengths and weaknesses
opportunity sampling

A

strengths: convenient, fast, cheap
weaknesses: not representative or random - sampling bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

✅❌
self selected sampling

A

✅ willing participants, can target specific people, easy and fast
❌ response bias (similar people) time consuming

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

✅❌
random sampling (database)

A

✅ representational, equal chance of selection
❌ restricted access to databases, not as replicable, not everyone is in a database

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

✅❌
snowball sampling

A

✅ find specific individuals
❌ time consuming, share common traits do not representative

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

✅❌
stratified sampling (maths to work out subcategories)

A

✅ representative
❌ smaller samples and time consuming

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

order of harvard referencing

A

last name first name
date
title
publisher
publisher place
page number

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

bps guidelines for ethics (7)

A

consent
debrief
right to withdrawal
protection from harm/distress
deception
confidentiality
responsibility

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

✅❌
questionnaires

A

✅ reliable (replicable and standardised), easy to administer, anonymous - willing to reveal, focused
❌ social desirability bias, sample bias (certain types of people)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

social desirability bias definition

A

answering in a way we believe is socially acceptable or desirable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

demand characteristics definition and examples x2

A

clues given to participants about aim of study - causing them to change behaviour
examples: leading questions, overt observations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

✅❌
structured interviews

A

✅ comparable (predictable and rigid) easy to administer
❌ social desirability bias, restricted data collection

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

✅❌
semi/unstructured interviews

A

✅ more qualitative data and more detail, flexible
❌ required skilled interviewers, low inter rated reliability, harder to compare, more susceptible to interviewer bias (their expectations)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

✅❌
self report

A

✅ qualitative and quantitative data, easy analysis, convenient, ethical
❌ social desirability bias, low response rate, can be up to interpretation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

split half technique

A

measuring consistency by asking same question twice (flipped)
increases reliability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

test re-test

A

when participants take the same test again at a later date - consistent results = reliable

23
Q

importance of pilot studies

A
  • increase validity - measuring what it’s supposed to
  • identifies problems
24
Q

✅❌
rating scales

A

✅ replicable, more depth to answers, easy to score
❌ biased by social desirability, set responses

25
Q

when to use wilcoxon matched pairs

A

ordinal data and repeated measures

26
Q

answering 15 markers (4 steps)

A
  1. choose RF
  2. define the technique or concept
  3. justify in context (link to strengths)
  4. add own experience and link to context
27
Q

how to include own experience in 15 markers

A

include twice (or more)
1st mention - detailed hypothesis and relevant detail
2nd - reduce the hypothesis or aim and focus on procedural detail

28
Q

type 1 error

A

falsely accepting alternate hypothesis
false positive

29
Q

type 2 error

A

falsely rejecting alternate hypothesis
false negative

30
Q

degrees of freedom calculation

A

(number of rows-1) x (number of columns-1)

31
Q

when to use chi square

A

independent groups and nominal data

32
Q

when to use spearman’s rank

A

correlations and ordinal data

33
Q

when to use sign test

A

nominal data and repeated measures

34
Q

✅❌ structured observations

A

✅ naturalistic, high EV, less demand characteristics
❌ low control, hard to standardise, influence of external factors

35
Q

why use overt observation

A

more control, informed consent

36
Q

why use covert observation

A

won’t be subject to demand characteristics, more natural

37
Q

what is time sampling

A

researchers return to location and undertake observations at set intervals, less time but more representative

38
Q

what is event sampling

A

researchers focus on a type of behaviour and is more specific, less chance of behaviour of interest being missed

39
Q

how to improve reliability of observations

A

inter-rater - use multiple observers that are trained on a coding scheme

40
Q

how to improve validity of observations

A

covert - ecological
several observations at diff times

41
Q

✅❌ repeated measures

A

✅ less participants needed, participants variables restricted
❌ order effects such as boredom

42
Q

✅❌ independent measures

A

✅ no order effects, less demand characteristics as unlikely to guess aim
❌ participant variables and more participants needed

43
Q

✅❌ matched pairs

A

✅ less participant variables, less demand characteristics and order effects
❌ more participants needed and harder to obtain

44
Q

✅❌ correlation studies

A

✅ predictive validity - can predict pattern and shows strength and direction of relationship
❌ limited uses as correlation not causation, can be reductionist

45
Q

what is hypothesis testing

A

making predictions about the likely outcomes of a study using objective methods
- can be null or alternate

46
Q

what is manipulation of variables

A

to test hypotheses in a valid and reliable way, where the IV and DV must both be defined and operationalised

47
Q

extraneous variables definition

A

anything external
1. situational variables such as noise, lighting, instructions
2. participant - intelligence, eyesight, boredom
3. experimenter - may give unintentional cues
4. demand characteristics - figure out aim, change to conform

48
Q

induction definition

A
  • start with data
  • conclusions from data
  • often qualitative
49
Q

deductive research definition

A
  • start with theory
  • confirm hypothesis with data
  • often quantitative
50
Q

quantifiable measurements definition

A

be able to quantify to make sure the dv is objectively measured

51
Q

positive vs negative skew

A

positive - high left, tail right
negative - tail left, high right

52
Q

what are structured observations

A

has a coding scheme with categories, coding frames are the abbreviations

53
Q

how to work out expected values (chi)

A

(row total x column total) / table total

54
Q

what are expected values

A

the answers you would expect to see if the null hypothesis is true