IPR Tests Summary PDF
Document Details
Uploaded by FinerVector3326
Durham University
Tags
Summary
This document provides a summary of statistical tests, focusing on chi-squared and correlational methods. It details assumptions, calculations, and interpretations for different types of data. It covers both parametric and non-parametric approaches.
Full Transcript
IPR summary of tests Chi squared test Assumptions for Chi squared 1. Nominal or ordinal data 2. Variables must be mutually exclusive/independent 3. Expected frequency should be 5 or more to ensure validity 4. Independent observations 5. Random sample Calculating Chi squa...
IPR summary of tests Chi squared test Assumptions for Chi squared 1. Nominal or ordinal data 2. Variables must be mutually exclusive/independent 3. Expected frequency should be 5 or more to ensure validity 4. Independent observations 5. Random sample Calculating Chi squared on JASP Chi squared goodness of fit test (one variable) 1. Calculate the descriptives 2. Frequencies 3. Multinomial test 4. Descriptives and descriptive plots When the frequencies have been calculated: 1. * Make sure the headings are correct 2. Frequency, contingency tables Chi squared test of association (two variables) 1. Calculate the descriptives- split one of the variables 2. Frequencies 3. Classical contingency table 4. Cells-count-expected When the frequencies have been calculated: 1. *Make sure the headings are correct 2. Frequency, contingency tables Results The bigger the chi squared value, the greater the difference between the observed and expected frequencies Power- 80 to reduce the chance of making a type II error P value-.05 to reduce the chances of making a type I error Correlational tests Parametric: data meets the assumptions (e.g Pearson’s r correlation) Non- parametric test: assumptions are violated (Spearman’s rank) Pearson’s r The direction and strength of a relationship between 2 variables and the covariance - Linear relationship Spearman’s rho The direction and strength of a relationship between 2 variables and the covariance - Non-linear (monotonic relationship) Calculating Pearson’s r and Spearman’s rho on JASP 1. Look at descriptives 2. Shapiro Wilk test of normality- if normal, continue with the analysis- far from 0.5 means that the distribution plot is normal 3. Basic plot-> distribution plot 4. Customisable plot -> scatter plot After descriptives... 1. Regression 2. Classical correlation 3. Display pairwise
