Every UX designer must be able to handle statistics and needs to do or at least understand some basic calculations.
Every UX designer needs to understand and calculate some basic statistics. The moment we evaluate significant amounts of testing results or our users' behavior on a website, we need to do some necessary calculations. I'm your host, Chris, and this is the UX Globals Podcast!
Don't get today's episode wrong. I do not say that we need to be experts in statistics. However, we need to understand what simple results like the modus, median, mean, and range are. The average is simply the number that the average person answered. Therefore, you add all numbers of participants answered to that one question and divide the result by the number of participants. If three participants answered nine and two answered eight. The sum is 43, and the number of participants is five. So the mean is 8.6. However, this number is not as good as the median of statistics as I will explain now. Let's say one more test was made, and this participant answeres only one. That makes a total of 44 with six participants, and therefore the result is 7.3. So it is way lower than the previous mean value, but it is only due to one participant. The median is created another way. You sort all results you have ascending. In our case it is 8, 8, 9, 9 and 9. Then you take the number in the middle, which is 9 in this case. If we add the 1 at the beginning the new numbers are 1, 8, 8, 9, 9 and 9. There is no number in the middle so we take the two numbers in the middle add them and divide them by two. That is (8 + 9) / 2 which results in 8.5. The difference between the two median is only .5 whearas the difference between the two mean values is 1.3. That sample shows that the median is often the better choice as it shows a more realisitc result compared to the mean value.
Let's address the modus at the end of today's episode. This number can also be very interesting to us as UX designers because it shows the number that was chosen the most by our user test participants. Let's retake our example. The numbers were one time a 1, two times an 8, and three times a 9. Therefore the modus of this sample is the 9 because it was taken three times.
Why do these numbers and statistics in general, matter to us as UX designers? Well, the moment we have to evaluate a system, we can not always choose qualitative feedback. If we make user tests with like one hundret participants, it will get very complicated to evaluate qualitative feedback. Therefore, we can select the quantitative questionnaires. The numbers entered by the users in the surveys need to be evaluated properly to properly adjust the tested solution. Due to that, every UX designer needs to understand what these metrics are and how we can use them to our advantage to improve the software, product, or service.
Join the UX Globals Community on Facebook, and remember to always keep pushing! Have a good one!