Inputs for chi-squared test

Hi.

I would like to test the hypothesis that the answers I got in a questionnaire from two differently sized patient groups have the same mean. The answers can be '1','2','3','4' or '5'. I was thinking of performing a chi-squared test, but am not familiar with the Statistics Toolbox and therefore cannot understand how to format my data as suitable inputs for chi2gof, crosstabs or any other of the Matlab functions that performs chi2 tests. Can you help?

Many thanks. Best wishes, Marta

Answer 1

[Aditya]

Hi Marta,

To test whether the distributions of responses from two different patient groups are the same, you can use a chi-squared test for independence. This test will help determine if there is a significant association between the group membership and the responses.

1. Organize Your Data: Create a contingency table that summarizes the frequency of each response in each group.
2. Perform the Chi-Squared Test: Use the crosstab function to create a contingency table and the chi2gof function to perform the chi-squared test.

% Sample data
% Group A responses
groupA = [1, 2, 2, 3, 4, 5, 3, 2, 1, 5, 4, 3, 3, 2, 1];
% Group B responses
groupB = [2, 3, 3, 4, 5, 5, 4, 3, 3, 2, 1, 5, 4, 3, 2, 1, 1];

% Combine data into a single vector and create a group label vector
responses = [groupA, groupB];
groups = [ones(1, length(groupA)), 2 * ones(1, length(groupB))];

% Create a contingency table
[tbl, chi2stat, pValue] = crosstab(groups, responses);

% Display the contingency table
disp('Contingency Table:');
disp(tbl);

% Perform Chi-Squared Test
% The chi2stat and pValue are already calculated by crosstab
disp(['Chi-Squared Statistic: ', num2str(chi2stat)]);
disp(['p-Value: ', num2str(pValue)]);

% Interpret the result
if pValue < 0.05
    disp('There is a significant difference between the groups.');
else
    disp('There is no significant difference between the groups.');
end

Make sure your data is correctly formatted and that the responses are categorical as expected by the test. Adjust the sample data to fit your actual dataset.