How the scores work
What the scores mean
5.0 Index scores run from 0 to 10, where 0 is always the worst possible score and 10 is the best possible score. A score of 5.0 represents the national average and serves as a benchmark. So, by glancing at the index score it is possible to gauge instantly whether it is above or below the national average.
How the traffic light system works
We use a traffic light system along with the 0-10 index to make the data easier to understand. It allows us to present your results in colours as well as numbers.
The traffic lights work with the index scores to illustrate very poor, poor, average, good and very good scores. We set the cut offs for each colour at approximately 20% so that each band represents about a fifth of respondents. This makes it easy to see if scores fall in the bottom 20% or the top 20%. The boundaries for the traffic lights are set out in the table below:
|Index score above||Index score below||Approximate % of respondents|
Note: due to the different distributions the % of respondents in each traffic light category is not precisely 20% for every domain, sub-domain and question. Across all the questions the overall variation is between 15-25%. We decided to keep the same cut-offs for all the traffic lights rather than introduce the extra complexity of letting them vary between questions and domains. It is best, therefore, to consider them as guides.
How we calculate the index scores
All questions give respondents 7 response options, labelled 1 – 7 for ease of use. Each scale is anchored at each end with positive (always 7) and negative (always 1) response options. Each question has a unique pattern of responses that can be described by their different means, standard deviations and skews. To aid ease of use these response patterns are ‘converted’ to 0-10 index scores.
It should be noted that this transformation is for communication purposes and is not suitable for technical analysis such as regressions and correlations. These should be carried out with raw results. Please contact us for details of our bespoke analysis services.
How we calculate the benchmark scores
The survey benchmarks are based on nationally representative surveys of 1,392 and 2,959 employed adults (aged 18 and older) in the UK and US respectively. The surveys used the same questions in both the UK and US and were conducted online at the same time in October 2011. The national benchmarks can be considered to be statistically accurate to within a 2% margin of error.
An ‘international’ benchmark is provided for non-UK & US residents. This should be considered as a ‘synthetic’ benchmark as it has been calculated using the full sample of 4,351 employed adults in the UK and US. Alternative methodologies were explored, including using the World Value Survey or European Social Survey as scaling devices. However using such techniques would undermine the advantage of our methodology, which is directly based on actual responses to questions in the survey. In line with this principle other national benchmarks will only be added when we have robust data for them.
Question response percentages
When answering the individual survey questions respondents use a 1 – 7 scale – where 1 always represents a negative response to the question and 7 a positive response. As another way of viewing the data we show the % of respondents scoring a low, medium and high score. For these purposes a score of 1-3 is categorised as LOW, 4 or 5 MEDIUM and a 6 or 7 as HIGH. They are shown as round traffic lights – with the size of the circles visually representing the percentage of respondents in these categories and the colours of red, amber and green representing the low, medium and high categories respectfully.
As these scores are actual responses they do not include the national benchmark information, however they can be used as the basis for internal organisational comparisons through time – i.e., in order to observe improvements in happiness at work by the rising percentage of employees in the HIGH category.