AVOID #2 - Validity & reliability: ask 100 people. Or more.

When you want to measure something, there are two important statistical aspects.’ the Transformation Executive continued. ‘Validity tells you whether your measurement is accurate. Reliability tells you how constant that accuracy is when you ask different people.’

‘When you do interviews, you mainly have a conversation.

There is not a fixed list of questions you’re going down.’

‘Exactly. Again, it’s good for getting a sense of direction but not for precisely measuring the transformation’s progress.

          ‘How does that work with verifiable questions like you showed

           me? Would it be different there?’ he asked.

‘When you have questionnaires that have been tested through the years, and many respondents did the test, the validity and reliability are not an issue. The respondents who answered the questionnaires before you were the guinea pigs which helped solidify the questionnaire's expected outcomes. They did the validity and reliability testing for you.’

          ‘Stan showed me how his framework is unique for every client.’

‘Correct. Every transformation has a different topic or accent. So, every questionnaire to employees is a one-off. No history to compare it with. Scientific research shows that you preferably have to ask 100 people or more for one-off, verifiable questionnaires. Then, the usual statistical measures for validity and reliability are in the green. The fewer people you ask than those 100, the less valid and reliable the outcomes are.’