# Types of research; descriptive

?
• Created by: becky_99
• Created on: 09-12-19 01:12

## Hypotheses

Thomas, Nelson & Silverman, 2014:

Hypothesis - "hypothesis deduced from theory or induced from empirical studies that is based upon logical reasoning and predicts an outcome of the study".

Null hypothesis - "hypothesis used primarily in the statistical test for the reliability of the results that says that there are no differences/relationships among variables".

1 of 6

## Variables

Independent variable - the thing that the researcher is manipulating, e.g. practice schedule.

Dependent variable - what the researcher is measuring, e.g. VO2 max.

Categorical variable - the things participants already possess, e.g. height.

Control variable - a factor that is deliberately kept out of the study, e.g. gender.

Extraneous variable - the things that may effect results, but kept in, e.g. age.

2 of 6

## Data classification

Non-parametric:

• Nominal scale - scale based on categories alone
• Ordinal scale - scale based on rankings

Parametric:

• Interval scale - scale which has equal units, but no meaningful zero
• Ratio scale - scale that has an equal distance between points, and zero represents an absence of that variable
3 of 6

## Descriptive research

Descriptive research provides data about the population being studied; cannot establish a cause-effect relationship between variables because there is no active manipulation of variables by the researcher.

It is also called observational because there is no intervention by the researcher.

4 of 6

## Types of descriptive research

• Cross-sectional studies
• Longitudinal studies
• Surveys (3 sub-types - questionnaires, interviews, normative surveys)

Questionnaires: closed and open-ended questions.

Correlational:

• Investigates whether relationships among variables exist; their direction and strength
• Alternative to experiments when not ethical/feasible to conduct
• Useful to predict variables using simple techniques
5 of 6

## Correlational vs. causation

Correlational is necessary but not sufficient condition for causation.

If no association between two variables, then no causation.

A significant correlational does not prove causation because there is no active manipulation of variables.

This causes:

• Third variable problem
• Directionality problem
6 of 6