Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

RM + Cognitive > Research Methods B.1 > Flashcards

Research Methods B.1 Flashcards

1
Q

When planning a histogram, what must we consider?

A

how many bins to use - increasing bins will give higher resolution, but too many can make the histogram too noisy

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

What is P-hacking, Publication bias, data dredging and HARKing?

A
  • P-hacking: manipulating analysis to get desired p-value
  • Publication bias - outcome of the experiment influences decision to punish or not
  • Data dredging - testing many hypotheses and only reporting the significant ones
  • HARKing - hypothesising after results are known
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What level of data is there no relationship between different possibilities in the scale?

A

Nominal level

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

What level of data is there a natural order between possibilities in the scale, but no interpretable magnitude of differences between values of the scale?

A

Ordinal level

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

What level of data are the possibilities ordered and have interpretable magnitudes, but zero doesn’t have a special meaning?

A

Interval level

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

What is the difference between discrete and continuous data?

A

Discrete data has clear spaces between values and is countable, continuous data falls in a constant sequence and is measurable

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

What is a test statistic?

A

a value that quantifies how close the data is to the null hypothesis

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

What is a P-value?

A

a value that shows the probability that we would have observed this same data if the null hypothesis was true

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the standard error of the mean and how can it be calculated?

A
  • the value that demonstrates the standard deviation of sampling distributions
  • calculated by dividing the standard deviation of the data by the square root of the number of samples
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

if SEM is too difficult to interpret, what is a better, more interpretable metric for the precision of our estimate of the mean?

A

confidence intervals: for example if we have 95% confidence intervals that mean we have a range of values which have a 95% chance of containing the population mean

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

In a normal distribution, what percentage of participants lie within one, two and three standard deviations?

A
  • 68.2% lie within one deviation
  • 95.4% lie within two deviations
  • 99.6% lie within three deviations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What test can be used to test if the data is normally distributed?

A

Shapiro-Wilk test

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

What do shapiro wilk W and shapiro wilk P tell us?

A

-Shapiro-Wilk W is a metric indicating how normal the data is, higher values = more normal data (if the value is greater than 0.05, its normally distributed)
- Shapiro-Wilk P is a probability indicating how significance any difference from normality is

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

What is ecological validity

A

the extent to which the variables and conclusions of a study sufficiently reflect the real world context of its population

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

What does a one-sampled hypothesis discuss?

A

if the mean of a particular data set is different to a specified value found ahead of time

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

What are the two reasons that we use a null hypothesis?

A
  • The null should be what we’re willing to assume until the alternate hypothesis is substantially proven (like innocent until proven guilty)
  • Its simpler: there are many ways that one thing can affect another, but only one way that there can be no effect
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

As the difference between the observed data mean and the comparison value gets bigger, what happens to the t-value?

A

it gets bigger

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

When reporting t-tests, what are all the values we need to include? (6 points)

A

Mean, SD, comparison level, t-value (with degrees of freedom in brackets) and p-value

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

What is the difference between within and between subject designs?

A

Between subjects: two independent groups: participants change BETWEEN conditions
Within subjects: each participants completes both conditions and contributes two data points

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

How do we calculate an INDEPENDENT samples t-test?

A

calculate the difference between the two means of the two groups of data, all divided by the pooled standard error of that difference (this is a single number that represents the variability of both groups, but it does assume that we have HoV because it assumes both groups have the same SD

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

What do different t-values indicate?

A
  • Large NEGATIVE t-value means mean of group one is less than group 2
  • Large POSITIVE t-value means mean of group 2 is less than group 1
  • near ZERO t-value means the means are close the the same
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What does a significant value from a levene’s test indicate, and what value would mean it was significant?

A
  • p values GREATER than 0.05 suggests there is no significant difference between variances, meaning we DO have homogeneity of variance
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

If our data fails the levene’s test and we don’t have homogeneity of variance, what alternative to a original t-test do we use?

A

Welch’s t-test, which uses an unpooled standard error of the difference

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

How and when do we used a paired sample t-test instead of an independent sample t-test?

A
  • when we use dependent groups
  • we calculate the difference between the pair of samples from the same participant in each condition to get a mean of paired differences, then divide that by the standard error of the mean paired difference
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

What does shapiro-wilk test test for, and if our data fails the test what do we calculate instead?

A
  • tests for normal distribution
  • if our data is not normally distributed, (value less than 0,05) we consider are non-parametric alternative such as wilcoxon (one sample test) or mann-whitney U (two sample test)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Why can’t we conduct a t-test on ordinal level data?

A

Because even though we could use the mean, t-tests also require the SD, which is hard to calculate with ordinal data because we don’t know the exact ‘distance’ between values on the scale

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What type of t-test do we use when comparing one sample to a reference

A

one sample t-test

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

What is an effect size?

A

it is a quantitative measure of the difference or relationship between two groups or variables

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

Why do we use degrees of freedom instead of number of observations?

A

because we have to account for number of values we’re estimating from the data
- One sample t-test DF = N-1
- Independent sample t-test DF: N1 + N2 - 2
- Paired sample t-test DF = N-1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

How do we report p-values?

A

Specify the degrees of freedom of the test, report exact p values to two or three decimals (but report p values less than 0.01 as < 0.01), specify the significance threshold used

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

what measure of effect size is usually used?

A

Cohen’s D

31
Q

What is considered a large effect size and what is considered a small effect size?

A

small is 0.2, medium is 0.5 large is 0.8
the larger the effect size the more meaningful the relationship or different between variables

32
Q

what do parametric tests assume?

A

that our data is normally distributed

33
Q

When do we use Wilcoxon signed rank test? (Non- parametric test)

A

when we have a one sample test or a dependent samples test

34
Q

When do we use Mann Whitney U test? (Non- parametric test)

A

independent samples design

35
Q

Why can Shapiro-wilk test not be the most robust test for normality, and what should we do to combat this?

A

because if our sample is too small, SW test may say its normally distributed even when on a graph its clearly not. and if our sample is too large, a SW test will be overly sensitive and will detect tiny departures from normality that we shouldn’t be worried about.
- to combat this, we should use multiple sources such as SW test, Histograms and QQ plots

36
Q

When do we use analysis of variance (ANOVA)?

A

when we are comparing MORE than two means/ when the independent variable has three or more levels

37
Q

What do we use to calculate ANOVA for the null hypothesis?

A

sum of squared deviances to the mean (sum square total)

38
Q

What do we use to calculate ANOVA alternative hypothesis for within groups design?

A

we use sum of squared deviances from the means WITHIN each group (sum square within)

39
Q

What do we use to calculate ANOVA alternative hypothesis for between groups design?

A

we use the sum square total minus the sum square within (sum square between)

40
Q

What is the final step in ANOVA calculation?

A

divide the SS within/between by either no. ppts - no. groups (within subject) or no. groups - 1 (between subjects)

41
Q

What does an ANOVA test tell us?

A

the total variability, and how much of this occurs within each group and how much occurs between. if there is a lot of variability between groups compared to within groups ( F will be large) then we have detects an overall difference/ significant effect, which we can explore further using post-hoc tests

42
Q

What are the assumptions of ANOVA for between subjects? (5)

A
  • interdependence (data is unrelated)
  • normally distributed
  • equal variance
  • categorical factors ( predicting factors must be divided into separate groups
  • data type interval or ratio
    assumptions all very similar to t-tests
43
Q

What are examples of traditional and alternative data collection methods?

A

Traditional: methods usually used within qualitative research e.g. interviews
Alternative methods: methods that are new or not regularly used e.g. qualitative surveys

44
Q

What are focus groups and what are they used for?

A

groups of usually 3-6 ppts where they involve sharing of experiences, ideas and views. Used to find out about ppts understandings and meanings, but from more than one person

45
Q

What is the difference between Asynchronous online focus groups and Synchronous online focus groups?

A
  • Asynchronous: conducted over a period of time, not all ppts need to be present at the same time
  • Synchronous: real time discussions usually done over video, basically the same as in person focus groups
46
Q

What perspectives do surveys usually suit?

A

realist, critical realist or essentialist perspectives

47
Q

What are ‘prompt methods’?

A

using videos/activities/ audios to start a discussion on a given topic

48
Q

What are some strengths of prompt methods? (2)

A

good for sensitive topics, and helps discussion become more participant led

49
Q

What are participant led prompts, such as photovoice?

A

giving ppts a camera to record their day to day life and make their lives accessible to others. It encourages the sharing of personal experiences

50
Q

According to Seitz and Orsini (2022) what are some key issues with photovoice papers? (3 points)

A

-the implementation of the method is inconsistent e.g. poor training of on camera use
-the implementation of evaluation of result is inconsistent e.g. measuring how empowered ppts felt
-the adherence to ethical procedures was inconsistent e.g. not gaining/ reporting ethical approval

51
Q

What is story completion?

A

where ppts complete a story stem usually started by the researcher. explore a range of assumptions of a given phenomenon and is a way of overcoming awareness of ppts own emotions

52
Q

What is a strength of story completion (1) and what are some things that we need consider?

A

-strength: useful for exploring socially sensitive or ambiguous issues
- must consider instructions, detail (not too much or too little), first or third person)

53
Q

What are solicited diaries?

A

Diary writing with pre-defined guidelines, e.g. travel practices or food consumption diaries (includes mobile app diaries and voice notes)

54
Q

What are some strengths of solicited diaries? (2 + 1 specific)

A

-(partial) access to thoughts/ feelings of ppts
-can feel cathartic for ppts as it provides them with a voice
- has been shown to be useful for caregivers of children with special needs as it improves holistic information

55
Q

What is media data and what is it useful for?

A

-using newspapers, magazines, tv and reader comments (easily accessible)
-useful for highlighting common messages and populations/issues

56
Q

What is user-generated content/ data harvesting?

A

-using pre- existing naturalistic data from online sources such as forums, blogs and social media
- Data harvesting = using forums/chats/tweets and analysing that text

57
Q

What is a strength of user generated content/ data harvesting?

A

its a good way of exploring phenomena without asking for it (useful for exploring interaction, group norms and linguistic patterns)

58
Q

What is the process of user-generated content? (4)

A

-select forum
-select the threads and the appropriate number of them
-download format ready for analysis
-select which elements of the data you need to focus on to answer your research question

59
Q

In user-generated content, what are some ethics you need to consider? (3)

A

-avoiding private discussion forums that you need to sign up to access
-considering if you need to ask an owner’s permission
- considering if you can make someone recognisable by quoting them

60
Q

What is thematic analysis?

A

a qualitative method of looking at data to find themes and identify patterns

61
Q

What is the different between the positivist, contextualist, and social constructionist view?

A
  • Positivism: human experience is knowable, universal and object. Knowledge is impartial, and research is an investigation for the TRUTH
    -Social constructionism: knowledge is historically and culturally contextualised and is (re) constructed through language. Research is an investigation of AN ACCOUNT of the truth
  • Contextualism: in between, believes there is no single reality, and believes that although the truth is inaccessible, knowledge can still be truthful
62
Q

What are the three ontological approaches?

A

-Realism: a pre-social reality exists that we can access through research
-Critical realism: the pre-social reality exists but we can only ever partially know it
- Relativism: ‘reality’ is dependent on the way we come to know it - there is no one reality

63
Q

What have been some problematic uses of thematic analysis?

A
  • a mashing together of approaches that don’t really suit each other
  • the use of coding reliability methods that seek out accuracy have undermined the embracing of subjectivity, which is more appropriate for quantitative methods
    -Misunderstanding of codes, themes and subthemes
64
Q

What is reflexive thematic analysis/ big Q research?

A

-research is active and embedded in results, reflecting on you position and your judgement in decision making. Focus on philosophy and procedure, rather than tools and techniques

65
Q

What is Interpretative Phenomenological Analysis (IPA)?

A
  • understanding participants’ experiences from their perspectives through use of interviews and small samples
66
Q

IPA is explained by the double hermeneutic, what is this?

A
  • First hermeneutic - ppts making sense of their experiences
  • Second hermeneutic - researcher making sense of the ppts’ sense making
67
Q

What are some differences between thematic analysis and interpretative phenomenological analysis?

A
  • IPA has clear philosophical assumtpions (critical realism, hermeneutics) wheras TA is more flexible to the researcher’s position
  • TA focuses less on the individual and more across the whole data set
  • TA can be applied to larger, more varied samples as the interpretation tends to be less in-depth/ less focussed on personal meaning
  • IPA is less focused on broader social structures
68
Q

What is discourse analysis?

A

Analysing and interpreting what people say and how they say it, and the cultural factors that influence communication. Heavily social constructionist and relativist, as its interested in how people use language to construct their own reality. NOT concerned with thoughts or feelings of participants

69
Q

What are some differences between discourse analysis and thematic analysis?

A
  • DA is associated with set philosophical assumptions (e.g. social constructionism, relativism) TA is more flexible
    -DA is more influenced by theory, whereas TA follows are more practise-based approach
70
Q

What is a type 1 error?

A
  • when our test gives a significant result when it shouldn’t: FALSE POSITIVE
  • to lower this a lower (more stringent) p-value should be used
71
Q

What is a type 2 error?

A

-When we accept the null hypothesis, even though its false. FALSE NEGATIVE
- To fix this, we cannot adjust it statistically, but we can improve by changing things in the experiment such as sample size

72
Q

What is meta-science?

A

using the scientific methodology to study the scientific process itself. Useful to identify strengths and weaknesses of the process

73
Q

What is reproducibility?

A

Using the SAME analysis on the SAME data and getting the same results

74
Q

What is replicability?

A

using the SAME analysis on DIFFERENT data and get similar answers?

75
Q

What is robust data?

A

using a DIFFERENT analysis on the SAME data and getting similar or identical answers. Checks whether a result is over-dependent on specific analysis software or choices

76
Q

What is generalisability?

A

the toughest standard: can we run DIFFERENT analysis choices on DIFFERENT data and get the same general result? This is what we’re aiming for.

RM + Cognitive (6 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions