A note on changing standard driver questions
Making changes to the predefined questions that Peakon uses can have adverse effects on the accuracy of benchmarking – which is relied on throughout the analysis Peakon provides. Even small changes can lead individuals to answer differently and potentially skew the results in comparison to the original questions.
Caution should be taken when editing questions, and when adding your own custom driver or sub-driver questions. Asking unclear, leading, or loaded questions could return a wildly different range of answers than those given to Peakon’s original questions.
Take caution when altering questions to be more specific
We’re sometimes asked about editing sub-driver questions to make them more specific. For example, changing the question “I am satisfied with the performance review process at our organization” to “I am satisfied with our process of running OKRs”.
While it may seem beneficial to include the name of a policy or methodology, this is a good example of how being more specific can pose problems. In this example, people may instead rate their impression of how OKRs are implemented, based on their knowledge of how OKRs should be practiced – rather than how satisfied they are with performance management in general.
If in a year’s time you were to switch away from OKRs to another methodology you will not have clean data to make an accurate decision about whether people are more satisfied. Instead, you’ll potentially have two entirely separate and unrelated reviews of how people thought processes were implemented, but on one trend line.
Peakon's smart text analysis applied to employee comments will help you surface the specific, recurring issues in your company, (such as OKRs in this example).
Best practice on formulating driver questions
Taking the following question as an example:
“I feel like my coworkers trust me to do a good job.”
We can be reasonably sure that feeling your coworkers trust you is positive, so this will fit the scale of 1 to 10.
But let’s have a look at ways this question could have been formulated that will cause problems:
“I feel like my coworkers don’t trust me to do a good job.” – this would invert the 1 to 10 scale, with 1 representing the positive, making a mess of the correlation to the engagement question.
“My coworkers should have more trust in me to do a good job.” – again, this would invert the 1 to 10 scale, as agreeing that you should be trusted more, indicates low trust.
“This is the most trusting team I’ve ever worked in.” – rather than gauging how trusting your team is, this would instead be entirely relative, and based on an employee’s previous career experiences.
“We have a world-class level of trust among employees.” – we have designed all of Peakon’s questions so that scoring 10 reflects excellence. However, setting such high standards as “world-class” would cause two problems: it’s essentially a higher standard than the scale of the other questions, and it’s unrealistic for employees to know what a truly world-class standard of trust is.
“Trust in our team has increased since our last team-building event” – time-sensitive questions are not a good idea – considering that you’ll need to ask the same question multiple times to understand whether improvements are being made. As time passes from the event in question, individuals’ memories of the event and its consequences will fade and differ, negatively affecting the reliability of their answers.
“Trust is essential for teamwork, I believe we have a trusting team.” – the framing of this question is not neutral. It insists employees agree with the premise before passing their judgment.
Hopefully, these examples will provide a useful guide for what to look out for when adding custom questions.
Best practice on formulating open-ended questions
Open-ended questions are a great way to collect feedback and ideas, which are not required to fit a 1-10 scale. These types of questions can easily be added or removed from your question rotation without affecting your scores down the line.
One good use of this feature would be to target offboarding and onboarding questions to new starters and leavers, respectively. This would only target those employees whose tenure and separation date details fall within the targeted time range. To read more about this, see Asking onboarding or off-boarding questions.
Open-ended questions are also a great way to gather and address questions from your employees in a Q&A dialog during company-wide meetings.
For example, having one open-ended question that asks employees what they would like to ask a senior member of the leadership team. The question could be similar to this:
Leadership can then incorporate into the regular company-wide meetings a short Q&A session that addresses some of these questions.