Crosstabulation table with column proportions tests
The crosstabulation table shows the percentages of employees in each job satisfaction category for each category of years with their current employer. For employees in the Less than 5 category, the percentages decrease with increased job satisfaction. To determine if the differences in those percentages, and thus the observed downward trend, are significant we'll examine the results of the column proportions tests.
The column proportions test assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions (for each row) are compared using a z test. If a pair of values is significantly different, the values have different subscript letters assigned to them.
For example, for the Less than 5 category, the value in the Highly dissatisfied cell has the subscript a and the value in the Somewhat dissatisfied cell has the subscript b. This means that the proportions in these cells are significantly different. In fact, each cell in the Less than 5 category has a different subscript, so we can conclude that the percentages are all significantly different. This also means that the observed downward trend in percentages of employees from the Less than 5 category with increased job satisfaction is real. Similarly, as you move to higher levels of job satisfaction you find larger percentages of employees from the More than 15 category. Since each cell in the More than 15 category has a different subscript, you can conclude that this upward trend is real.
Values in the 5 to 15 category do not show a consistent trend with increasing job satisfaction. In addition, the values in the Somewhat dissatisfied, Neutral and Somewhat satisfied categories all have the same subscript b, which means they cannot be distinguished from each other. However, the subscripts in the Highly dissatisfied and Highly satisfied categories (for the 5 to 15 category) differ from each other as well as from the common subscript for the other three categories. This means that the percentages in the Highly dissatisfied and Highly satisfied categories are significantly different from each other as well as from the percentages in the other categories.