In order for statistics to provide an accurate picture of the reality managers must act upon, two basic requirements must be met: 

• The questions must be a clear and precise operationalisation of the areas we want to explore and research
• When using a scale from 1-6 and wanting to calculate meaningful averages, we must ensure a smooth transition between the scores while also justifying the number of alternatives

The operationalisation of the questions is about the precise linguistic design of the questions when aiming to explore conditions at work.  How questions are phrased is of great importance to ensure we measure what we think we are measuring – also across departments and over time. The questions have therefore been formulated so that they do not require advance knowledge and clearly focus exclusively on one thing at a time. Care has also been taken to avoid loaded or leading questions.

At times we have chosen to incorporate questions that linguistically encourage a yes/no answer. However, given the context of the scale from 1-6, we believe that all employees immediately will realise that a graduated response is desired. It must be emphasised that the questions illicit each employee’s subjective evaluation of their workplace. However, this is naturally of objective value for the manager, who can, for example, observe whether there is generally a high (or low) level of satisfaction in one or more areas, based on the averages which are automatically generated by the system.

In order to calculate an average, one must methodologically use either interval-scaled variables, or variables that approximate interval-scaled variables. The latter are used in our forms here. Employees are able to provide responses on a scale from 1-6, but we allow the system to calculate an average that lies between these discreet values (e.g. 4.3). We do this because we are confident that the responses could, in principle, be infinitely graduated. But this has to be weighed against usability in practice.

We have chosen not to include a neutral median (such as ‘3’ on a scale from 1-5). This is because we assume there generally is a trend one way or the other for each employee, which is what we are interested in. Our scale will force this to be expressed, even in cases where the employee might wish to answer neutrally for the sake of politeness or convenience. Our selection of a scale from 1-6 is the result of many years of experience and experimentation. A scale with few choices and no median would force employees to always take a strong position, whereas a scale with too many choices would not be fully utilised.

We does not include instructions in the forms indicating what a top or bottom score should represent. We have avoided this because the form aims to measure each employee’s subjective experience of the given factor. It is up to each manager and the director to use their judgement and statistical awareness to decide whether it is reasonable to use the results from Musskema.dk as a basis for comparisons over time and between different departments. For example, if there are radically different cultures in departments to be compared, or if virtually all the staff in a department have recently been replaced, there may be cause to question whether the comparisons are reasonable. However, under normal circumstances, the results from Musskema.dk can be seen as very reliable.

If you have any comments or questions about the above, you are very welcome to contact Developandperform.com via email (info@developandperform.com).