# Last minute revision for Stats exam

Assumptions for an ANCOVA
Same as ANOVA but we have linear regression and homogeneity of regression coefficients
1 of 61
What is the assumptions of linear regression?
Deviations from the regression equation across the different levels of the independent variable must have normal distributions with the means of zero and homoscedasisity
2 of 61
What is homogeneity of regression coefficients
The regression coefficient for each of the groups in the IV should be the same
3 of 61
When is a MANOVA conducted?
ANOVA style analysis on several DV variables
4 of 61
What are the assumptions?
Multivariate normality: Sampling distributions of the DV and all linear combinations of them are normal, Homogeneity of variance, Boxs M test this is but it is advised that p.80
5 of 61
singularity
A variable is redundant; a combination of two or more of the other variables
6 of 61
The aim of discriminant function analysis?
Find a set of variables that predict membership of groups
7 of 61
When is it used?
When groups are already known and the researcher is trying to find out what the differences are between the groups
8 of 61
Kant
Duty or right based Ethics
9 of 61
Bentham
outcome based ethics
10 of 61
3 main codes of ethics?
Harm to human pps must be avoided, informed consent, debriefing
11 of 61
What are the three Rs?
Refinement, reduction and Replacement
12 of 61
Internal validity?
The extent to which the results obtained are a function of the variables that were systematically manipulated
13 of 61
Maximise internal validity
Rule out the possibility of other factors producing the change, must control everything and eliminate exteneous influences, easiest in highly controlled, lab settings
14 of 61
History effects
Events occuring during the experiment that are not part of the treatment
15 of 61
Diffusion of treatment effects
Occurs in control group learns about the manipulation
16 of 61
Compensatory Rivalry
When pps in different conditions start competing
17 of 61
Compensatory Equalisation
What if the experimenter knows what group the pps is in
18 of 61
Resentful demoralisation
If the control group learns that they are in the control group they may stop trying
19 of 61
Population validity
The extent to which the results can be generalised from the experimental sample to a defined population
20 of 61
Ecological validity
The extent to which the results can be generalised from the set of environmental conditions in the experiment to other environmental conditions
21 of 61
Reactive effects of testing
Occurs when a pre test increases or decreases the respondents sensitivity to the treatment
22 of 61
Reactive effects of experimental arrangements
These can occur when the conditions of the study are such that the results are not likely to be replicated in non experimental situations
23 of 61
Selection treatment interaction
The possibility that some characteristic of pps selected for the study interacts with some aspect of the treatment
24 of 61
Multiple treatment inference
When pps receive more than one treatment, the effects of previous treatment may influence subsequent ones
25 of 61
Replication
An additional scientific study that is conducted in exactly the same manner as the original research project
26 of 61
Pearsons
Interval + ratio and Interval + ratio
27 of 61
Spearman's Rho:
Ordinal/ rank and ordinal/ rank
28 of 61
Kendall's Tau
Ordinal (Rank) + ordinal (Rank)
29 of 61
Phi
True dichotomy x2
30 of 61
Structured interview?
Questions and response categories that predetermined by researcher
31 of 61
Unstructured
List of topics to discuss with pps; interview strongly pps led
32 of 61
Semi structured
REsearcher has a list of questions but scope for pps to discuss issues important to them and not anticipated by rresearcher
33 of 61
Convienience sampling
Selection most accessibly pps
34 of 61
Purposive sampling
Generate insight and in depth understanding of the topic of interest
35 of 61
theme
A patterned response in your data that has a meaning in relation to the research question
36 of 61
Six phases of thematic analysis?
Familiarising yourself with data, generating inital codes, searching for themes, rev
37 of 61
Semantic code
Succint summary of explicit data
38 of 61
Latent content
Provide an interpretation about the data
39 of 61
Split half reliability
Items are split into two halves
40 of 61
Parallel forms
Create a large pool of items, randomly divide the items into two seperate tests
41 of 61
Cronbach's Alpha
Mathematically equivalent to the average of all possible split half estimates
42 of 61
What does a figure of +0.7 indicate?
indicates acceptable internal reliability
43 of 61
What is the Kuder Richardson Formula
Measures internal reliability for measures with 2 choices
44 of 61
External reliability
Test retest reliability, perform the same survey with the same respondents at different points in time
45 of 61
Inter-rater reliability
Determines the estent to which two or more raters obtain the same result when coding the same response
46 of 61
Validity
If a test is valid then you measure what you intend to measure
47 of 61
Content validity
The extent to which a measure represents all faces of the phenomena being measured
48 of 61
Construct validity
Seeks to establish a clear relationship between the construct at a theoreritcal levela dn the measure that has been developed
49 of 61
Convergent validity
The measure shows association with measures that it should be related
50 of 61
Discriminant validity
The measure is not related to things that it should not be related to
51 of 61
Predictive
Assess whether a measure can accurately predict future behaviour
52 of 61
A Priori
Consider the proposition, "If George V reigned at least four days, then he reigned more than three days." This is something that one knows a priori, because it expresses a statement that one can derive by reason alone
53 of 61
A posteriori
Compare this with the proposition expressed by the sentence, "George V reigned from 1910 to 1936." This is something that (if true) one must come to know a posteriori, because it expresses an empirical fact unknowable by reason alone
54 of 61
What counts as a moderate skew?
1.96 to 2.33
55 of 61
What counts as a substantial skew?
2.34 to 2.56
56 of 61
What counts as a severe skew?
2.56 >
57 of 61
Bartlett's Test of Sphericity
Identity matrix: Variables are unrelated
58 of 61
Kaiser Meyer Olkin Test
Proportion of variance might be caused by underlying factors
59 of 61
How do you work out an observed score in questionnaires?
True Score + Error
60 of 61
What is reliability related to?
Average correlation between items and test length
61 of 61

## Other cards in this set

### Card 2

#### Front

What is the assumptions of linear regression?

#### Back

Deviations from the regression equation across the different levels of the independent variable must have normal distributions with the means of zero and homoscedasisity

### Card 3

#### Front

What is homogeneity of regression coefficients

### Card 4

#### Front

When is a MANOVA conducted?

### Card 5

#### Front

What are the assumptions?