Research Design and Appropriate Analyses

Question 1

If the dependent variables (outcome variables) are categorical (nominal/ordinal), and there is only ONE IV, which analysis is appropriate to use?

Answer 1

Logistic Regression (if the predictor variable is continuous) or Pearson chi-square (if the predictor variable is categorical)

Question 2

If the dependent variables (outcome variables) are categorical (nominal/ordinal), and there are TWO IVs, which analysis is appropriate to use?

Answer 2

Logistic Regressions or analyses

Question 3

If the dependent variables (outcome variables) are CONTINUOUS (interval/ratio) data, and there are TWO or more IV’s, which analyses would be appropriate to use?

Answer 3

• ANOVA’s
  Either Factorial Independent/Factorial Repeated Measures/Factorial Mixed Design

OR

• Multiple Regression/ANCOVA

Question 4

If the dependent variables (outcome variables) are CONTINUOUS (interval/ratio) data, and there are TWO or more IV’s that are both continous, which analyses would be appropriate to use?

Answer 4

Multiple Regression

Question 5

If the dependent variables (outcome variables) are CONTINUOUS (interval/ratio) data, and there are TWO or more IV’s that are both categorical, which analyses would be appropriate to use?

Answer 5

ANOVA -

Factorial Independent/Factorial Repeated Measurs/Factorial Mixed design

Question 6

If the dependent variables (outcome variables) are CONTINUOUS (interval/ratio) data, and there are TWO or more IV’s that are both categorical and continuous, which analyses would be appropriate to use?

Answer 6

• Multiple regression/ANCOVA

Question 7

If the dependent variable (outcome variables) are CONTINUOUS (interval/ratio) data, and there is only ONE IV that is continuous, which analyses would be appropriate to use?

Answer 7

Pearson correlation or Linear Regression

Question 8

If the dependent variable (outcome variables) are CONTINUOUS (interval/ratio) data, and there are TWO IV’s that are categorical, which analyses would be appropriate to use?

Answer 8

Either
Independent T Test (if independent or between groups) or
Dependent T Test (if repeated measures/within groups).

Question 9

If the dependent variable (outcome variables) are CONTINUOUS (interval/ratio) data, and there are >2 IV’s that are categorical, which analyses would be appropriate to use?

Answer 9

Either
One Way Independent ANOVA (if independent groups)
or
One Way Repeated Measures ANOVA (if dependent groups)

Question 10

In hypothesis testing, what does the null hypothesis tell us?

Answer 10

There is no relationship or there are NO group differences

Question 11

In hypothesis testing, what does the alternative hypothesis tell us?

Answer 11

There is a relationship or there ARE group differences

Question 12

In statistical tests/analyses, what are test statistics and what do they tell us?

Answer 12

Test statistics are values that describe the DEGREE to which the relationship between variables OR group differences vary from the null hypothesis of no difference or no relationship

Question 13

In statistical tests/analyses, what are p-values and what do they tell us?

Answer 13

P-values estimate the likelihood in which one would see the difference described by the test statistic if the null hypothesis were true

Question 14

In statistical tests/analyses, how do we decide which one to use?

Answer 14

1. What data is used and does it meet assumptions for parametric tests in general?
2. What types of variables do we have? Categorical or continous?
3. Is the research design repeated measurements of independent observations?
4. How many predictors and outcome variables? Number of levels in each variable?
5. Once test selected, check if data meets assumptons for that specific test.

Question 15

Does Nominal Data have ordering to values?

Answer 15

NO. Has no ordering to gender, sex, eye colour, etc. It’s arbritrary

Question 16

Does Ordinal data have logical order?

Answer 16

YES. Grade, birth order, race result