# PSYA01H3 Chapter Notes -Statistical Inference, Naturalistic Observation, Statistical Significance

16 views1 pages

PSY Exam Prep – Chap 2

5 Steps in the Scientific Method:

•Question

•Method

•Collect

•Analyze

•Publish

Types of studies:

•Naturalistic observation: observe behaviour of people in their natural environment

•Correlation study: observe and see if 2 variables have a relationship. The magnitude can

range from correlation coefficient -1 to +1 (either negative correlation: one goes up one

goes down, or negative correlation: both go up).

•Experiment: change one variable and see if it effects another variable

Nominal fallacy: the incorrect belief that one has explained a behaviour just by describing it.

Operational definitions:

In order for an experiment to be accurately duplicated and produce the same results each time,

accurate descriptions of the variables must be provided. These are called Operational

Definitions: defining a variable in terms of the operations a researcher used to

measure/manipulate it.

If the operational definitions are correct, then the experiment has Validity.

Researchers must remain alert to the problem of confounding participant characteristics with

their independent variable manipulations.

•Ex: if a researcher wants to see how anger affects concentration, he would have to make

some participants angry. They might exert their right to leave the experiment at any time,

and now the control and experiment groups are not equivalent anymore.

Correlational study example: are shyness and daydreaming related?

Generalization occurs in 2 ways:

•We generalize observations from our sample to a different group

•We generalize from one context to another

Descriptive statistics: (what are the results?)

•Measures of central tendency: mean, median, mode

•Measures of variability: range and standard deviation

Inferential statistics: (are the results significant?)

•Statistical significance: does a relation exist or is it due to chance?

•Inferential statistics: mathematical procedures tell us if the relations are statistically

significant