Am I running this Chi-Square right on Julius?


I just wanted to make sure I was running this correctly on Julius:

Below is the contingency table from a survey testing to see if there is any relationship between job satisfaction and education level.

Education level Dissatisfied Neutral Satisfied
College 16 37 64
High School 25 26 41
University 25 21 45

According to the results displayed, there was no association between educational level and job satisfaction (χ²(2) = 8.4486, p = 0.076). Did I display this right?

Thanks in advance!

1 Like

Hi Emily,

Yes! Perfect! Good job! Just remember to check your dataset and make sure that it follows the correct assumptions. They are as followed:

  1. Observations should be independent:
    This means that each observation in the dataset should not be influenced by any of the other observations. It looks like you are asking people to rate their satisfaction you probably safely assume that their feelings are strictly their own and should not influence their peers’. But, also double check to make sure.

  2. Random Sampling
    The survey should have been distributed in a way that was random. Random sampling is very important in the integrity of results because it ensures that the samples are representative of the population from which they were drawn (avoids biases). So as long as the prompt says it used random sampling to distribute the survey, then we can say that it also passes this assumption.

  3. Categorical data
    In order for the chi-square test to be run, categorical variables must be used. You seem to have different education levels, as well as satisfaction levels. Those are both considered categorical variables because we can group or categorize them, so we can also say that it passes that assumption.

  4. Expected cell frequencies
    The frequency count for each cell in the contingency table must be at least 5. When expected frequencies fall below this number, the test may not provide accurate conclusions. It looks like your lowest value in the contingency table is 16, so we also pass this assumption.

Hope this helps! And, again, great job!


Thank you for the kind reply, and also further information on the chi-square test! Appreciate it!