20 Questions
What is the primary assumption of independence of observations in chi-square tests?
That each participant must fall in one and only one category.
What is the minimum expected cell frequency required for a chi-square test?
5
What is the consequence of having small expected frequencies in a chi-square test?
The test becomes more lenient.
Why is it essential to include non-occurrences in chi-square computations?
To ensure that all participants in the sample are accounted for.
What is the purpose of special forms of tests that exist for repeated measures designs?
To test for relationships in repeated measures designs.
What is the implication of having a chi-square value of 1.20 that is not significant?
There is no significant difference between groups.
What is the primary advantage of using a chi-square test?
It is a non-parametric test.
What is the consequence of violating the assumption of independence of observations in a chi-square test?
The test is invalid.
What is the purpose of checking the expected frequencies in a chi-square test?
To ensure that the expected frequencies are at least 5.
Why is it essential to include all participants in the sample in chi-square computations?
To ensure that all participants are accounted for.
What does a moderate level of association between two variables imply?
A moderate to high level of correlation between the variables
What is the purpose of using Fei Square as an index of variance explained?
To explain the variance in categorical data
What is the effect of squaring the correlation coefficient value of 0.5?
The correlation coefficient value becomes 50.35
What is the relationship between class attendance and passing or failing an exam according to the correlation analysis?
Class attendance has a moderate correlation with passing or failing the exam
Why are parametric tests, such as Pearson's correlation, considered more powerful than non-parametric tests?
Because they are more robust to assumptions
What is the purpose of using point-biserial correlation?
To analyze the relationship between a continuous and a categorical variable
What is the implication of a chi-square value of 50.592?
There is a moderate correlation between the variables
Why is it essential to check the expected frequencies in a chi-square test?
To ensure the assumptions of the test are met
What is the purpose of using chi-square tests for categorical data?
To analyze the relationship between two categorical variables
What is the advantage of using parametric tests over non-parametric tests?
They are more robust to assumptions
Study Notes
Chi-Square Assumptions
- A participant must fall in one and only one category, and cannot be counted twice.
- Observations must be independent, meaning chi-square tests are not suitable for repeated measures designs.
Expected Frequencies
- All expected cell frequencies should be at least 5; if not, the chi-square test should not be used.
- Small expected frequencies produce few possible values of chi-square, which are compared to a continuous distribution.
- The greater the degrees of freedom, the more lenient this requirement is.
Inclusion of Non-Occurrences
- Computations must be based on all participants in the sample.
- Inclusion of non-occurrences is essential, as demonstrated in the mirror self-recognition study example.
Example: Mirror Self-Recognition Study
- A study with 40 2-year-olds (20 boys and 20 girls) tests gender differences in mirror self-recognition.
- The observed results are: 17 girls pass the test, and 11 boys pass the test.
- The expected results are: 14 girls and 14 boys would pass the test.
- The chi-square value is 1.20, which is not significant (ns).
Assumptions of Chi Square Test
- Independence of observations: each participant can only fall in one category and cannot be counted twice
- Size of expected frequencies: expected cell frequency should be at least 5; if not, Chi Square test should not be used
- Inclusion of non-occurrences: computations must be based on all participants in the sample, including those who did not pass the test
Limitations of Chi Square Test
- Restricted to less than 22 or less variables to test; for more variables, Multivariate generalization of Chi Square approaches is needed
- Sample size plays a strong role in significant testing; if the sample size is doubled, the obtained Chi Square value also doubles
Effect Size of Chi Square
- Chi Square does not indicate the strength of a relationship; a significant result does not necessarily mean a large effect
- Phi (φ) is used to measure the effect size of Chi Square; it can be interpreted as the correlation between two dichotomous variables
- Phi can be interpreted in the same terms as Cohen's conventions for effect sizes
Interpreting Phi
- Phi = 0.5 indicates a moderate level of association between two variables
- Phi = 0.65 indicates a moderate to high level of association between two variables
- Phi Square can be used as an index of variance explained, but it's not actual variance since there's no variance in categories
This quiz covers the assumptions of Chi-Square tests in psychology, including the importance of participant categorization and independence of observations.
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free
