statistics: 5.2 (PARAMETER) ASSUMPTION CHECKS

how to choose a statistical test? 3 pts

it depends on two things:

1. the type of data you have (nominal, ordinal, quantitative)

2. your research question- regarding relationship between variables and differences between groups or conditions

3. parametric vs non-parametric

parametric vs nonparametric tests meaning? 2 pts

1. parametric- makes assumptions about a population’s parameters e.g. mean or standard deviation

2. nonparametric- does not assume anything about the underlying distribution; you know the population data does not have a normal distribution

requirements of parametric tests? 5 pts

1. quantitative variables (interval or ratio)

2. normally distributed data

3. no outliers

4. equal variances

5. larger samples

requirements of nonparametric tests? 5 pts

1. qualitative data e.g yes/no

2. abnormally distributed data

3. outliers present

4. unequal variances

5. small samples

basic requirements and assumptions: binomial tests?

1. assumption 1→ a variable can be nominal or ordinal but must involve only two response variables e.g. yes/no male/female pass/fail

2. assumption 2→ the trials in which participants respond are independent

3. assumption 3→ representative sample of the population through the use of simple random sampling

basic requirements and assumptions: chi-squared test? 2 pts

1. assumption 1→ the variables can be nominal or ordinal but must be categorical

2. assumption 2→ each variable must comprise of at least two categories

types of t tests? 3 pts

1. One-sample t-test compares a mean value to a predetermined value

2. independent samples t-test- compares two means that come from different groups

3. paired samples t-test- compares two means that come from the same group

basic requirements and assumptions: one sample t-test? 4 pts

1. assumption 1→ the dependent variable must be quantitative

2. assumption 2→ each observation must be independent from the other; one participant’s score should not influence anothers

3. assumption 3→ dataset should not contain any outliers

4. assumption 4→ dependent variable should be approx normally distributed in the population

Basic Requirements and Assumptions: independent samples t-test? 6 pts

1. assumption 1→ dependent variable must be numerical

2. assumption 2→ dependent variable should be approximately normally distributed within each group

3. assumption 3→ independent variable must be categorical with exactly independent two groups

4. assumption 4→ each ppt must only belong to one group and observations should be independent

5. assumption 5→ data should not contain significant outliers

6. assumption 6→ variability of scores should be similar in both groups

Basic Requirements and Assumptions: paired samples t-test? 4 pts

1. assumption 1→ dependent variable must be numerical

2. assumption 2→ dependent variable should be approximately normally distributed within each group

3. assumption 3→ two sets of scores must come from the same participants or matched participants

4. assumption 4→ data should not contain significant outliers