1. Research Principles Flashcards
What are some types of epistemological frameworks and what type of methodology do they line up with?
Positivism/Realism = quantitative/ scientific method
Contextualism/ Phenomenology/ Constructionism = qualitative
What form of measurement should you go for?
Go for the highest form of measurement where possible (numeric over ordinal over nominal)
What are the 3 types of categorical data?
Binary (2 categories)
Nominal (more than two categories)
Ordinal (logical rank order: fail, pass, distinction or in a race)
What are the 2 types of continuous data?
Interval (equal intervals e.g. 6 & 8 = the difference between 2 & 4)
Ratio (e.g. a score of 16 is double a score of 8- true zero point)
What are the different types of sampling you can use with observation studies?
Event (when event occurs)
Time (fixed intervals)
What are some advantages to fully structured observational studies?
Systematic
Controlled for confounds
Replication
Constant setting
What are some advantages to naturalistic observational studies?
Ecological validity
Rich
Less demanding setting
What are some disadvantages to fully structured observational studies?
Lack ecological validity
Not spontaneous
Ethical issues
What are some disadvantages to naturalistic observational studies?
Observer obtrusive
No control over confounds
Difficult to replicate
Ethical issues
What is an example of secondary data?
Census data, Websites, Clinical and police records, Organisational records, Newspapers and magazines, Letters and diaries, Research articles
What type of methodology would you use if you were using secondary data?
meta-analysis,
qualitative meta-synthesis
What does a case study examine?
A single unit of analysis e.g. an individual, organisation, group, intervention
A case study can either be ?
Descriptive or Explanatory
What are the 3 types of interviews?
Semi-structured
Structured
Unstructured
What do focus groups aim to capture?
Understandings, perspectives, stories, discourses and experiences not otherwise meaningfully expressed by numbers
What are the disadvantages of a focus group?
Requires moderator have good facilitation skills.
Rich data but may be difficult to analyse.
Group processes affect data obtained. Artificial setting
What are some different types of questionnaires?
open-ended closed-ended categorical multiple-response ratings (e.g. likert) rankings
What is problematic when using questionnaires?
The assumption that psychological or emotional quantities can be measured using interval scale
What are quasi-experiments?
Experiments where participants can’t be randomly assigned to groups
What are some examples of qualitative data collection?
Semi-structured interviews Life story, episodic or narrative interviews Case studies Autobiographical memories Focus groups
What are some examples of qualitative analyses?
GT IPA DA FDA Narrative MW
What type of analysis would you use for comparing 2 groups of different participants?
Independent t-tests
What type of analysis would you use for comparing 3 or more groups of different participants?
One way ANOVA
What type of analysis would you use for comparing more than one grouping factor?
Factorial ANOVA
What type of analysis would you use for comparing 3 or more groups of the same participants?
Repeated measures ANOVA
What type of analysis would you use for comparing more than one repeated measure factor?
Factorial repeated measures ANOVA
What type of analysis would you use for comparing both independent and repeated measures groups?
Mixed factorial ANOVA
What type of analysis would you use for comparing one or more numeric variables?
Correlation (Multiple regression)
What type of analysis would you use for comparing both independent groups and numeric variables?
ANCOVA
What types (4) of analysis would you use for comparing differences?
t-tests
ANOVA
ANCOVA
MANOVA
What types (6) of analysis would you use for comparing relationships?
Correlation Linear Regression Meditational regression Logistic regression Path Analysis SEM
What is the shapiro wilk?
A correlation between the obtained distribution of scores and what the scores should be if they are distributed normally
When is it best to use the Shapiro Wilk statistic?
In smaller samples
What do we use to measure the normal distribution statistic in larger samples?
Kolmogorov-Smirnov
What could you do (but isn’t recommended) if you have moderate positive skew?
Square root transformation
What could you do (but isn’t recommended) if you have severe positive skew?
Logarithmic or inverse transformation
What could you do (but isn’t recommended) if you have extreme positive skew?
Fourth root transformation
What could you do (but isn’t recommended) if you have negative skew?
Reflex transformation (subtract each score from max score + 1)
When is bootstrap a useful alternative to non-parametric estimates?
When normality is violated
When homogeneity of variance is violated
When a non-parametric test is not available
Which models don’t have equivalent non-parametric tests?
Regression models
What are the possible strategies for dealing with outliers?
Remove (bad)
Trim (okay)
Winsorizing (good)
What are the 2 rules for trimming the data?
percentage (e.g. remove 5% from each end) standard deviation (e.g. greater than 2SD in z scores)
What are the 3 types of missing data?
MNAR (bad - dependent)
MAR (okay- 1/2 dependent)
MCAR (good - independent)
What are the strategies for dealing with missing data?
Delete the case (listwise - good - least biased)
Estimate (inflated SE = less power)
Expectation Maximisation (EM - done on totals)
What are the 2 types of Estimating Missing data?
Mean substitution (BAD) Regression Substitution
What is the statistic that determines the independence of the missing data?
Little’s test (should be non sig.)
What should you do if you find a sig. result from Little’s test?
Listwise deletion
