Question | Answer |
When to use non-parametric statistics? | very severe violation of assumptions, DV is not interval or ratio level. |
Logic of hypothesis testing? | state the null
adopt an alpha level
test the null
reject or fail to reject the null. |
TWO ways that researchers can control for the effects of extraneous variables or turn them into control variables? | build the variable as a factor--sample the control for them by limiting the sample.
OR
Randomly assign and do ANCOVA. |
List the criteria for evaluating quantitative methods. | generalizability
control of error
minimize threats to internal and external valitidy
replicability |
Describe data interpretation approaches in single subject designs | visual inspection
statistics |
3 types of mixed methods. | simultaneous.
quan then qual.
qual then quan. |
Difference between true experiment and quasiexperiment | both have control of IV.
true is RANDOM assignment.
quasi is NOT random assighment |
Describe general procedures for organizing and “reducing” qualitative data | gather data, code transcribed data for categories, identify themes, determine the pervasiveness of themes, interconnect themes |
List several of the “units” of focus for qualitative research | roles, relationships, groups, organizations, habitats, subculture |
Types of qualitative sampling | representative, purposive (e.g., maximal variation), snowball |
Identify factors which influence the necessary sample size for a study | number of groups, desired power and alpha, desired effect size, |
CIPP evaluation model | Context (when where)
Input (what is going into the program)
Process (how does it happen)
Product (outcome--what do we get out of it) |
Practical significance | Effect Size needs to be above---0.25 to be considered practically significant |
Statistically significany | needs to be below .05 or .01. |
What does the ALPHA value measure? | tells you how many times out of 100 is acceptable to be due to chance. More than that--not statistically significant. |
Phenomenology | telling a story (narrative) |
Ethnography | studying a culture, a group |
Positives to Random Selection | Have the populations, the group you wanna study.
Then selects the random group.
Allows a lot of external validity (our sample reflects our population) |
Positives to Ramdom Assignment | Once you have your sample
Randomly assigning them to 3 groups
Benefits: controlled for extraneous variables |
APA definition of Evidence Based Practice | combination between research and clinical judgement. (must have literature and support from clinicians) |
Null Hypothesis | Stating that there is NO relationship between what you are studying. |
Alternative Hypothesis | Thinking that treatment A is better than treatment B. You may actually test the null hypothesis, but this is what you actually think. |
Sampling Error | Statistical analysis account for this, the sample group doesn't EXACTLY represent the actual population data. |
Sampling Bias | NOT OK. THIS IS BAD. Sampling Bias is when you organize a group to get the results you want |
Sample vs Population | Sample is a smaller group to represent the population. |
Ways to fix violations of "assumptions of independence" | treat the group as an individual
lower the alpha value to .01
(multiply the p value by 4) |
Types of Quantitative Sampling | Random
Stratified
Cluster
Convience |
What is random quantitative sampling? | I know the population, I pull randomly from the population |
What is stratified quantitative sampling? | My sample has the same representations (example:% white, % AA %hispanic) as the whole population |
What is cluster quantitative sample? | Randomly sample groups (example: classes at UAB.) |
What is convenience quantitative sample? | I sample who is in easy for me to sample. CANT generalize this. BAD NEWS! |
Standard Deviation and Variance both tell you... | how far ON THE AVERAGE out from the mean. |
Standard Deviation and Variance are difference because | Variance is square. |
VALIDITY | |
RELIABILITY | |
Positively Skewwed | the mean is being pulled higher....due to outliers |
Negatively skewwed | the mean is being pulled lower...due to the outliers. |