W8: Data Analysis II Flashcards
What are tests of significance? Give examples.
to test hypotheses
eg. chi square, t-test, ANOVA
What is a research (alternative) hypothesis, Ha?
prediction of relationship between variables
What is a null hypothesis, Ho?
prediction that there is no relationship between variables (by convention, this is a statement of equality)
Which hypothesis is tested in a test of significance?
the null hypothesis is tested
What does it mean to accept the null hypothesis?
conclude that there is no difference (no relationship) between variables
What does it mean to reject the null hypothesis?
conclude that there probably is a difference (relationship) between the variables
what is another way to say “accept the null hypothesis”
fail to reject the null hypothesis
What is publication bias? What are the impacts?
occurs if results from studies which have not been published are different from those that are published
impacts interpretation of reviews and meta analyses that include only published literature; overestimation of effects of treatment
(analysis with multiple variables) Define the main effect.
differences among groups for a single IV that are significant, temporarily ignoring all other IVs
(analysis with multiple variables) define interaction effects.
differences among groups of a single IV that are predictable only by knowing the level of another IV
what is the difference between a one and two tailed test?
one tail: specifies direction of a relationship in advance (“directional” hypothesis)- if you predict direction of a relationship in advance, you do a one tailed test
two tailed test: test of any relationship between variables, regardless of direction of relationship
- if you do not predict direction of relationship, you do a two tailed test
what are sampling fluctuations?
how much stat value fluctuates from sample to sample.
Not reflective of true difference in population from which sample was selected
what is statistical significance? What type of error can occur?
degree of risk that you are willing to take that you will reject a null hypothesis when its actually true
-> concluding that there’s a stat. sig. difference exists when there is actually no difference
(type one error) = level of statistical significance
What is the difference between statistical and clinical significance?
clinical difference-> importance of a difference; decided by clinical expertise
stat significance-> says nothing about actual magnitude or importance of difference
What are the three tests of significance?
- Chi- square, X2
- T-tests
- ANOVA (extension of t-tests)
When is chi-square used? What types of data are the DV and IV? What is the null hypothesis?
used with cross tabular analyses
IV: categorical (nominal/ordinal)
DV: nominal level
Ho: no significant difference between categories on variable of interest
When is t-tests used? What types of data are the DV and IV? What is the null hypothesis?
compare means of 2 groups (means of DV); when sample size <30
DV: ratio level
IV: has two levels (categories) only
Ho: no significant difference between group means on variable of interest (dependent)
What is the difference: between and within t-tests
between subjects t-test (2 independent samples): use in experimental design, with experimental and control group
within subjects t-test (2 dependent samples): same person subjected to different treatments and comparison is made between two treatments
What is an analysis of variance (ANOVA)?
family of statistical tests- compares group means to assess whether differences across means are reliable
What is a post hoc comparison?
comparison of differences across levels of an independent variable when results are significant
What are the IV and DV in an ANOVA?
IV (treatment): 2+ categories (nominal/ordinal); can be more than 1 simultaneous IV
DV: measured at ratio level
When are tests of significance not appropriate?
- access to total population
-non-probability/random sampling procedures used - high non-participation rate
- non experimental (qualitative) research
- when not guided by formal hypothesis
- when data set is limited
What are inferential stats used for? What does it use for data analysis?
to determine probability that conclusion based on analysis of data from sample is true
- use tests of significance for data analysis
