data analysis, interpretation, and presentation

approach

depends on data

quantitative

qualitative

mixed methods

translating raw data

interviews

translating raw data

interviews

questionnaires

translating raw data

interviews

questionnaires

observations

analysis

begins with initial reactions or observations

identify patterns

calculating values

data cleansing: check for errors

interpretation

parallel with analysis

results interpreted different ways

make sure data supports conclusion

avoid biases

avoid over claiming

presentation

different methods, depends on goals

affects interpretation

quantitative analysis

statistical analysis

averages

percentages

averages

mean

median

mode

individual differences

task times: 45, 50, 55, 55, 60, 65

mean: 55

median: 55

mode: 55

task times: 45, 50, 55, 55, 60, 300

mean: 94.2

median: 55

mode: 55

task times: 10, 10, 50, 55, 60, 300

mean: 80.8

median: 52.5

mode: 10

data format

table: rows and columns

how to represent responses and measures

participants as rows

responses: single or multiple

bar charts

scatter plots

pie charts

presenting percentages

almost always better ways

comparison difficult when percentages similar

include percentages as text

provide overview with other representations

qualitative analysis

initial approach

gain overall understanding

look for interesting features

highlight common, record surprises

tools

transcription: oTranscribe, Otter.ai, Trint

coding and analysis: Dedoose, Atlas.ai, Nvivo

approaches

identifying themes

categorizing data

analyzing critical incidents

identifying themes

thematic analysis

inductive

identify, analyze, and report patterns

themes represent important, relevant, or unexpected patterns

steps

iterative

open coding: initial pass

axial coding: themes across participants, connections, categories

selective coding: validate relationships, consistency, themes

affinity diagrams

organize ideas and insights into hierarchy

groupings emerge through data

categorizing data

deductive vs inductive

study goals

deductive: categorization schemes

inductive: arise from analysis

Interface Problems
1. Verbalizations show evidence of dissatisfaction about an aspect of the interface.
2. Verbalizations show evidence of confusion/uncertainty about an aspect of the interface.
Content Problems
1. Verbalizations show evidence of dissatisfaction about aspects of the content of the electronic text.

questions?

reading for next class

Chapter 15: Evaluation Studies
Interaction Design: Beyond Human-Computer Interaction