Using Quantitative Techniques with Traditionally Qualitative Methods
Sometimes we need statistics, but most times, we dont. In this tutorial, Ill take you through several case studies where Ive strengthened the data that I gave to my clients by combining qualitative techniques and quantitative data, using simple math and graphs. Most times, you dont need statistics do prove a point, answer a question, or settle an argument.
But for those times when stats are really necessary, Ill take you through some basic statistical methods using nothing but Excel. If youre currently using confidence intervals and p-values, you should check out Jeff Sauros tutorial. But if you arent using basic stats and want to learn how and when you should use them, this is the place to start.
- User Research
- 9:00am to 5:15pm on Monday, June 04, 2012
About the Tutorial
GOALS FOR THE SESSION:
Participants will get new insight into how to combine qualitative and quantitative techniques, to prevent and respond to questions such as:
"That's not a statistically significant sample size, is it?"
"How do you know that you talked to the right six customers?"
"How can you generalize that finding across all of our users or across our personas?"
"If we make this change, what will the impact be?"
"How can we measure the improvement in the user experience, or the value of that change?"
"The preferences and behaviors of those two groups of participants are in conflict with one another ... how do we know which group to design for?"
"Which changes are the most important ones to make?"
In the morning, well ensure were all on the same page by defining what we mean by qualitative and quantitative data and methods. Then well move into case studies where quantitative data was used to prioritize usability study findings and to measure the value of proposed changes identified in heuristic evaluation. Well talk about validation, and then move into case studies where quantitative methods led to unanticipated benefits when conducting focus groups and creating personas.
After lunch, well review statistical terms such as descriptive, inferential, discrete, continuous, sample, mean, median, mode, normal distribution, and standard deviation. Doesnt that sound fun? No? To make it fun, well play some games, which will involve chocolate as incentives for participation. Then well talk about when and how to use significance tests and confidence intervals. After that, we might have some more chocolate, as we discuss p-values and Excel. The Excel exercises will be done as a class, and Ill be sure to give you resources on where to go for more information and learning. Well conclude the afternoon as we break into teams to for a data analysis case study.