Chi Square test

The the Chi Square test is used to investigate if distributions of categorical variables differs from one another (Ordinal Scale).
There are 3 different modes in the Chi Square test:

1. Two way count data
2. Equal proportions
3. Specific proportions

These can be selected with the check box.

Two way count data

To test the difference between to groups of categorical data.

Interpretation

• The smaller the p value is the more likely there is a significant difference between the 2 data-sets.
• Develve uses the commonly accepted value of p < 0.05 for significance.

Colors of the cells

• Green
  No significant difference
• Yellow
  Significant difference

Formula

Legend

counts
expected counts
Column
Row

Example

There is no significant difference between the data-sets in column D and E.

Data file

Equal proportions

To test if there is a difference between one of the category and the rest of the data.

Interpretation

• The smaller the p value is the more likely that at least one proportion in the data set is significant different.
• Develve uses the commonly accepted value of p < 0.05 for significance.

Colors of the cells

• Green
  No significant difference
• Yellow
  Significant difference

Formula

sum of counts / amount of options

Legend

counts
expected counts

Example

One of the counts in Column A is significantly different.

Data file

Specific proportions

To test if there is a difference between categorical data and the expected data.

Interpretation

• The smaller the p value is the more likely there is a significant difference between at least one of the proportion and expected value.
• Develve uses the commonly accepted value of p < 0.05 for significance.

Colors of the cells

• Green
  No significant difference
• Yellow
  Significant difference

Formula

sum of counts * percentage (in second column)

Legend

counts
expected counts

Example

There is no significant difference between Column A and the expected ratios.

Data file

External links

• http://en.wikipedia.org/wiki/Chi-squared_test