The Two Branches of Statistics How to Transform Observations into Variables Three Types of Variables Introduction to Hypothesis Testing
Sample
Population
Inferential
Descriptive
Descriptive
Inferential
Population
Sample
Because most researchers are lazy
Because it is usually impossible to study an entire population
Because populations can only be studied if the data is continuous
Because samples are typically more accurate than populations
Ordinal
Continuous
Discrete
Nominal
Ordinal
Interval
Nominal
Ratio
Nominal; ordinal
Interval; ratio
Ordinal; interval
Discrete; continuous
If our independent variable had too many levels
If our dependent variable was continuous
If we were using a between-groups research design
The presence of a confounding variable
Measures what it was intended to measure.
Predicts actual behavior.
Is consistent.
Co-varies with the independent variable.
Not valid.
Not reliable but probably valid.
Not reliable.
Neither reliable nor valid.
Operational definitions.
Hypothesis testing.
Specificity of measurement.
Continuous observations.
Operational definitions.
A correlational study.
Random assignment.
Random selection.
Between-groups
Within-groups
Correlation
Independent groups
Descriptive statistics.
Inferential statistics.
Nominal variables.
Ordinal variables.
Randomly assigning participants to conditions of an independent variable.
Drawing conclusions about the relationship between variables.
Using sample data to make general estimates about the larger population.
Organizing, summarizing, and communicating a group of numerical observations.
Independent variable.
Population.
Sample.
Dependent variable.
Continuous
Ordinal
Discrete
Nominal
Ratio
Interval
Ordinal
Nominal
Nominal variables.
Ordinal variables.
Interval variables.
Ratio variables.
The sample is small.
It rained the day of the physical test.
The company funding the test was biased.
People who are regular sports drinkers may be more athletic than those who are not.
Variables.
Operational definitions.
Levels or conditions.
Confounds.
Activity level of the child.
Type of puppet used by the experimenter.
Number of imitated actions.
Length of the experiment.
Involves each participant participating in all levels of the experiment.
Includes only one condition.
Is sometimes referred to as a longitudinal design.
Involves each participant experiencing only one level of the experiment.
Correlational
Repeated measures
Within-groups
Between-groups
Within-groups
Between-groups
Experimental
Correlational
Descriptive statistics.
A correlational research design.
A within-groups research design.
Inferential statistics.