EMPLOYEE SATISFACTION SURVEY > Tabulating Survey Data

Table of Contents

Tabulating Your Survey Information

One of the most important aspects of conducting an employee satisfaction survey is tabulating the results appropriately.  A scientific, statistical approach must be used to tabulate this information in order to achieve optimal results.  Collecting groups of random comments without proper categorizing is a sure recipe for disorganization and, ultimately, failure.  Issues must be coordinated among logical employee work groups in order to clearly plan how to attack problems that are identified (and even which problems to attack).  No matter which method of gathering opinion is utilized, it is necessary to organize the data in a way that is useful for strategy and planning purposes.

The first principle of organizing satisfaction survey data is to group comments into categories of opinion.  For example, comments regarding pay and benefits should be separated from comments regarding policies, procedures and other matters.  The table below lists the categories of behavior LRI Management Services assesses during employee satisfaction surveys.  Of course, categories can be added or subtracted based on the specific needs and issues of the company.

In addition to categorizing opinion according to question category, the results are also tabulated according to other applicable categories, i.e.:

By breaking survey results down according to these areas, we can pinpoint exactly where problems are occurring in the workforce and where they are not.  Failure to break results down into work group categories like these can often hide significant problems in the company.  Some groups may be more positive and mask the scores of groups that are more negative.  Pivoting the data in these various ways provides the clearest picture of the company and creates opportunities to attack issues that may have remained hidden under other methods.

There can be problems with breaking down data too narrowly.  If the size of the work unit is small, or if the number of breakdowns is too numerous, there is the real possibility that only a few employees will fit into any one category.  This creates two major problems.  First, it can reduce the reliability of data, due to the fact that employees may fear that management will be able to identify individual employees who made a comment.  This may result in employees answering untruthfully to statements about their supervisor or management. 

There are also legal issues raised by the National Labor Relations Board regarding surveys that identify the individual making comments (see page XX for more information).  Therefore, it is vital to reach a good balance between enough data to make strong recommendations and too much data, which can negatively impact the results of the survey.  Most of this planning should occur in the initial survey design.  Work with your LRI advisor to determine the most important categories to use for your organization.

Care should also be taken in determining how opinion is evaluated among the categories and work groupings because there are several different ways to look at issues.  For example, one might look at the overall average score in a particular work grouping.  Another angle is to compare the percentage of employees who responded favorably to those who responded unfavorably.  The results can also reveal specific numbers of employees who have rated statements at various levels of agreement or disagreement along the spectrum.  There are advantages and disadvantages to each type of rating, and all three should be used to best analyze issues in a company.

Overall Scores
Overall scores are good for gauging, in general terms, how a group feels about a particular statement or category.  These overall scores are misleading at times because extremely high ratings can sometimes mask a significant negative opinion on a question or issue.  For example, a department consisting of three shifts may have very strong favorable opinions on first and second shifts but extremely negative opinions on third shift.  While a significant problem exists on the third shift, it is not evident in the overall score.  As a result, a company may be unaware of a significant employee relations issue.

Percent Favorable to Percent Unfavorable
This method is effective for determining significant pockets of favorable or unfavorable opinion.  It allows a company to pinpoint where extremes have pulled a score up or down. 

However, percent favorable to unfavorable does not give as accurate a view of the overall feeling of a work grouping as the overall average score.  This is because the favorable and unfavorable ratings can be spread across various degrees of agreement or disagreement. 

For example, a company may have an evenly distributed number of favorable and unfavorable responses, as a percentage, but those responses may be unevenly distributed within the favorable and unfavorable categories.  See Appendix 2 for an example.  In other words, two companies with the same percent favorable to percent unfavorable groupings could have significantly different overall scores on the same issue.

Actual Response Distribution
The advantage of this method is that a picture is formed of how employees feel on particular issues across the entire spectrum of responses.  This method takes out the “averaging effect” of using overall favorable to unfavorable responses.  The possible disadvantage of this method is that it creates information overload in some respects.  In other words, negative opinion is expressed to many questions. Without some sense of the overall score of a particular group, it becomes increasingly difficult to prioritize action areas. 

Therefore, it is easy to see the importance of using all available information, broken down in a variety of ways, to determine exactly which areas deserve the most consideration in a particular company.  The charts and graphs in Appendix 2 show how your data will be presented.