I am working on a application which already had user research documents, now i am adding my user research document to it. At every usability session more and more documents been created. How i ensure at a point of time that i am not going a step back in usability

Several usability practitioner (owner) had worked on one interface. I will represent it with one example, assume initial system was 30% usable

First usability practitioner done related research after usability testing application was 50% usable sounds good.He left the job

Second usability practitioner done related research without looking into previous research carried out by First usability practitioner, after usability testing application was 35% usable, he left the job

Now Third usability practitioner who what to compare current usability testing results with the previous usability practitioners documents at a point of time in project to ensure that application usability remain step forward after each iteration

Two main reason of comparing

  1. To make sure that "application usability is increasing gradually in the project life cycle"
  2. To get the feedback early in the time for example - if usability of the application is decreasing you will immediately come to know. You will stop there rectify it thus company will not save time but also money
  • Can you provide a little more information? What are these documents, what do they contain and how are they being used?
    – Matt Obee
    Nov 8, 2013 at 9:14
  • Doing more research isn't ever going to be step back. I don't really understand what it is you have an issue with here - what is the link between research documentation and 'stepping back in usability'?
    – JonW
    Nov 8, 2013 at 10:14
  • 1
    I think this is an excellent question: particularly on an interface which has had a number of 'owners' over a period of time - owner 1 may have done research on design 1; drawn conclusions; implements changes; and then owner 2 comes along does research on design 2; draws different conclusions; implements changes; etc etc. So depending on the 'draw conclusions' stage it is quite possible for an interface to get worse over time not better.
    – PhillipW
    Nov 8, 2013 at 11:23
  • imo, the project/application needs an UI/UX design expert who can "crush" all those documents to bring out the juice, and who has a final say on design. looks like "too much non-cooks are spoiling the broth" ... :|
    – kmonsoor
    Nov 9, 2013 at 22:28
  • Since the system evolves over time as its requirements and implementation are changed, I think the real issue should be how to test usability of existing and new users-tasks over time and make sure that the new user-tasks are both usable and don't break usability of old user-tasks (assuming they haven't been removed). Nov 12, 2013 at 12:40

2 Answers 2


This is a case where there is a lot of miscommunication because you don't have clearly defined usability guidelines. Here is a recommended series of steps to take to ensure the guidelines are communicated and enforced by the usability practitioner.

  • First take all the user research documents from previous iterations
  • Define the pain points that users are facing
  • Define the funtionality and the expected use cases
  • Define the user guidelines and expected level of support (for example : If you are testing for accessibility clearly define the level of support i.e. Section 508,A,AA or AAA) so the tester will know what guidelines to follow.) You can also base your usability guidelines based upon existing standards e.g. Nielsons Usablity guidelines
  • Define your heuristics based upon that if the usability practitioner is going to do heuristic analysis as well
  • Work closely with the usability practitioner to ensure that he is aware of the guidelines to follow and how he defines if an application is usable or not.

"who want's to compare current usability testing results with the previous usability practitioners documents"

I'd read up on the previous research but wouldn't attempt to measure against it claiming any degree of precision:

Usability Issues are 'loaded' - ie

Issue One - Very Important

Issue Two - Moderate Significance

Issue Three - Showstopper

The 'loading' process is, at the end of the day, subjective, so any previous numerical total of issues x importance loadings is also going to be subjective.

So I'd concentrate on trying to improve what you've got rather than attaching much importance to comparison with previous results.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.