Testing of Assumptions: Statistical Research Guide

Summary

Statistical procedures rely on assumptions. This guide explains these assumptions, how to test them, and the consequences of violating them. It helps students understand key concepts in statistical research.

Common Assumptions

Many statistical procedures share common assumptions. These include sound measurement, avoiding attenuation, and avoiding tautological correlation. The guide explains each of these.

Normality and Skew

Most parametric tests assume normal distribution of data. Skew measures the tilt of a distribution. The guide describes how to test for normality and interpret skew.

Outliers and Multivariate Outliers

Outliers can distort statistical results. Simple outliers affect a single variable. Multivariate outliers affect multiple variables. The guide explains how to identify and handle them.

Multicollinearity

Multicollinearity occurs when independent variables are too highly correlated. It makes it hard to tell their individual effects. The guide explains how to detect and address multicollinearity.

Key terms

Parametric statistics

Statistics that assume a certain data distribution, usually normal, and interval-level measurement.

Skew

The tilt of a distribution, either to the right or left.

Outliers

Data points that are far from the rest of the data.

Multicollinearity

High intercorrelation among independent variables in a regression model.

Homogeneity of variances

The assumption that different samples have the same variance.

Box plot

A chart that shows the spread of data for different categories.

Questions this document helps with

• What are the common assumptions in statistical procedures?
• How can I test if my data meets the assumptions of a statistical test?
• What should I do if my data violates a statistical assumption?

Summary written with AI from the document's text.