The systems that our society uses to measure how people feel simply do not work at a certain level. This includes especially surveys used by various organisations to analyse people's sentiments.
The internal data gathered by companies provides a clear idea of whether employees feel committed and satisfied. But if executives happen to access references, for example, on portals where employees, both past and present, anonymously evaluate company culture, they are taken aback: in many cases external sources indicate employees are less engaged than internal surveys suggest. Sometimes, though, it's the other way around.
At first glance, it may seem that external sites must have skewed data, thus creating a false impression, because the people supplying the information may not be a valid sample of the employee base. However, Engelberg states his research in companies often reveals it is in fact the internal data that contains a number of deviations.
There are several reasons for this. Some of them reflect the problems that plague both political and various opinion polls. The sample may be unrepresentative; for instance, it is extremely difficult to determine whether those completing the internal surveys are representative of employees across the entire organisation.
Rajeev Peshawaria, author of Open Source Leadership, for example, suggests some employees are too busy to complete the surveys at all, while others deliberately avoid them because they do not believe their answers will be genuinely anonymous.
The employees who complete the internal surveys are in fact determined by self-selection, and this is an unrepresentative sample of participants.
At the same time, it is not always clear how each respondent interprets a given question. Another problem companies face is an inability to define issues clearly in order to prevent different interpretations. For example, asking about an employee's level of engagement based on a scale from 1 (lowest) to 10 (highest) may lead to distorted answers.
For instance, an employee who feels highly motivated and involved in tasks they enjoy will record a mark of 10. On the other hand, an employee who feels they enjoy a well-balanced work and personal life may decide on a 5. Thus people with similar levels of engagement will actually offer a completely different assessment in the survey.
Surveys can be designed to offer specific and clear definitions of terms, leaving room for extensive qualitative data in which people explain in their own words why they answered the way they did. Detailed information may also be included to clarify how anonymity will be ensured. Sometimes, however, data relevance issues persist. People are simply too complex, so the data provides only clues rather than insight.
So what does all this mean? That companies should rely less on data to understand employees and predict their behaviour. In a nutshell:
- Ask less, observe more. To find out how your employees feel and what challenges they face, look around and make a note what you see. Do people work well together? Do they treat each other with respect? Does everyone get a chance to talk and be heard?
- Avoid "pseudo voices". When employees point out problems, they often slip through cracks and have no impact. The so-called "pseudo voice" gives employees the "illusion of participatory influence". Unfortunately, this is also common at companies that claim to be committed to giving employees a chance to contribute their ideas. If you want to involve employees, take all feedback seriously.
- Look ahead. Focus more on building a better future and less on assessing the state of affairs right now.
-bb-