Review the Results
( First page > Data Analysis > Review Results)
These notes outline the steps in actually making sense of the data. The discussion presumes that you worked through the statistical analysis of the data; now we're looking for implications rather than just lots of numbers.

1. Check Key Outcome Measures
   1. Check the level
      1. These measures reflect the purpose of the study. Whether they are high or low will (should?) be the first point of concern. One of my clients had chosen an outcome measure capturing their hope for a work environment free from fear. They were shocked (appropriately so) to find the level so astonishingly low. The rest of the analysis orbited around that single fact.
      2. Remember that even if the levels are high, later we'll be asking what are the strongest contributors. If you're doing someone really well (i.e., it has produced high morale, optimism, or loyalty), then it's important to understand what has made the difference and reinforce the practice. For example, if the data analysis shows that managers' interpersonal skills (giving feedback, compassion, listening, etc.) were the main determinants of high loyalty or morale (a typical finding), then that might change the way you hire new managers or train and promote current managers.
   2. Review distributions
      1. Check for anomalous distributions. Even with a high overall scores, looking at the actual distributions may reveal a noticeable cluster at the lower end of the scale.
      2. Check for significant demographic differences. Are the outcome measures different by any of your demographics? Is morale (or executive credibility or faith in the strategic plan) the same across departments? Across tenure groups? Across job types? Where is it highest? Where is it lowest?
2. Check potential contributors
   1. Check the level
      1. Look for strong contributors that are relatively high in their level. These are the important things that you're doing right. Think about how those practices could be extended, reinforced, preserved.
      2. Look for strong contributors that are relatively low in their level. These are the important things with lots of room for improvement. Think about how to reverse the current practices and start off in a new direction.
      3. Be especially mindful of contributors that show up as having significant impact on 2 or more key outcome measures. These are the action areas with more bang for the buck.
   2. Review distributions
      1. Check for anomalous distributions, such as heavily skewed distributions, flat distributions, or even bimodal distributions.
      2. Check for significant demographic differences
3. Identify and rank order issues
   1. Which of the key outcome measures are especially strong?
   2. Which of the key outcome measures are particularly low?
   3. Which of the potential contributors show the greatest impact on the outcome measures?
   4. What kind of tone is revealed in the comments? Are people hopeful? Or apathetic?
   5. What additional issues surfaced in the comments that warrant additional exploration?

Back to Top