Mind your attitude!

May 21 2010 by Brian Amble Print This Article

Employee attitude surveys that rely heavily on agree/disagree scales can deliver misleading results that lull employers into false sense of security, according to one of the UK's leading surveying experts.

Many surveys are based almost entirely on lists of agree/disagree statements. This format makes it easier to ask questions on a lot of issues and is simple for respondents.

But some critical questions will simply don't fit into this format and it provides no means to explore what lies behind employees' answers.

For example, while an answer may reveal that an employee does not feel that their line manager does a good job of managing them, the issues that explain why are not revealed.

In addition, by only measuring attitudes, such surveys fail to consider knowledge, behaviour, and motivation which are all highly relevant to employee engagement levels.

Speaking at the Institute of Internal Communication's conference earlier this month, Peter Hutton, who heads up BrandEnergy Research and was former deputy Managing Director at polling giant MORI, said that agree/disagree statements should make up no more than 10% of an employee engagement survey.

"Any survey that genuinely aims to understand employees and help to identify appropriate future actions, will contain a variety of different question types that have been carefully designed with the specific needs of the organisation in mind," he explained.

"To support organisational objectives, the starting point needs to be your business model and the desired knowledge, behaviour and attitudes of staff - something that is definitely not achieved by a standardised survey.

"Clearly, having the right kind of bespoke questionnaire involves more work. This even extends to survey respondents, who have to think about different question types as they go through, but this in itself is a positive as it means that they will consider individual questions carefully and cannot go into automatic mode."

  Categories: