Management

OSCR and me: Anonymity

What does “anonymous” mean in the context of a stress audit which asks extremely personal questions? This article looks at how the information is collected, used and reported, and shows how your identity is protected.

The anonymity of the respondents is key to getting honest results, so how do we go about ensuring that anonymity, and what information does your employer get about you? How is the personal information used, and is this just a big excuse to to check up on you?

Password login

The first question asked by many is “How can the survey be anonymous when we’ve each got unique passwords?”

Why passwords?

We need to have some form of barrier to try to prevent somebody taking the survey twice and skewing the results (yes, the survey is so much fun, people want to do it again and again). The password is a low level, simple means of doing this. By having a unique password for each time the survey is accessed, we can limit it to the correct number of respondents.

Two databases

“Surely, you’ve sent me my password to my email account - you must know who I am?” Well, that much is true, but there are a few things to notice: first, the requirement for entry is not your email, merely a matching username and password. Secondly, we actually use two separate databases, one to authorise each user, the second for the survey data, so, at no point is your email address, username or password associated with your actual answers.

Information in the survey

Why ask for such personal information in the survey?

The first section of the survey asks detailed questions about you, such as your age, sex, department, role, length of service, absence days etc. Why do we ask for these? Some of these items are used for sorting, whilst others are used as indicators. Sorting questions are sex, age, department, role, etc. They allow us to sort the data into groups for analysing. The other questions, like medication, alcohol consumption, absence days etc are indicators which, along with other answers, lead to your overall assessment score.

Columns, not rows

If you are unique in your sorting answers, can you be identified? The short answer is that if you are the only 63 year old female manager in the sales department, then yes, you can be identified. However, there are two key things to remember, first that at this point it is off-site assessors that are looking at the data, i.e. people who do not know you, and secondly, that the assessment doesn’t look at you as an individual.

Imagine that we’ve got a big table of all the results, with each row being one person’s responses to the survey. The information we want to get out of this survey is all to do with the columns - how many males are stressed, how many managers are stressed, how many people have a high absence rate. We don’t look at the rows. If 10 people are identified as being in a high risk stress bracket, we’re looking at the columns to see if there are patterns. For instance, if 9 of the 10 are from one department, then we know there is an issue in that department. That can only come from looking at the columns, not the rows.

Information given to employer

The information that is given to your employer is a detailed report on overall trends within the group survey. We do not give the survey answers to the employer. Equally, we do not give any summary information which would lead to individuals being identified. For instance, we may summarise the breakdown in the sales department by gender and age, but the admin department by age only, because to use gender would highlight an individual. So, what information does your employer get about you? As a member of a group, whether that be a department, a gender or an age bracket, your results are contained within the average data, but as an individual, your results are never given to the employer.

Employer’s liability

From a legal and liability point of view, your employer really doesn’t want to know the identity of the individuals suffering from high levels of stress, and would rather deal with anonymous groups than individuals.

Yes, but no

So, to summarise, the information you provide to us in the survey can be used to identify you, but the key things to remember are:

• your identity isn’t attached to the data.
• the raw data is looked at by people who don’t know you.
• we look at the data in columns, not rows. We never look at indivdiual results, it’s always groups.
• your employer gets summary data, never individual results.

© 2008 stress and culture audits.