Lecture 4 - Surveys

Question 1

What do surveys allow?

Answer 1

allow capture of high level information about user experiances , attitudes and perceptions

Question 2

How can surveys be administered?

Answer 2

paper , phone , email , website

Question 3

Why are surveys good?

Answer 3

They’re low cost and have a large reach

Question 4

Are poorly designed surveys good?

Answer 4

No, they provide noise

waste of our and respondents time

Question 5

Are surveys themselves a topic of research?

Answer 5

YES
includes:
- population sampling
- optimising data collection (return rates)
- reduce biases in questions
- question order effects
- computer vs paper based (as surveys predate computers)

Question 6

What did we do in AE1-1?

Answer 6

We used heuristic analysis to identify a summary of issues and how visible they are. It doesn’t however label issue seriousness.

Question 7

Are well designed surveys good?

Answer 7

Yes, they can provide valuable insight.

Question 8

When can surveys be used in a project or research?

Answer 8

Can be used pretty much at any stage for a range of purposes.

Question 9

what are the 4 cognitive steps to survey responses?

Answer 9

• comprehension of the question , instructions and answer options
• retrieval of specific memories to aid with answering the question
• judgement of the retrieved information and its applicability to the question
• mapping of judgement onto answer options

Question 10

What questions are surveys good for?

Answer 10

• attitudes
• intent
• task success
• feedback on UX (did you enjoy it)
• user characteristics (demographic , gender)
• interactions with technologies
• awareness
• comparisons
• it may be a good idea to survey regularly to assess changes over time

Question 11

What is an interrupt question?

Answer 11

One that is within the middle of the task in order to avoid having to recall!

-> people not great at remembering

Question 12

What questions are bad for a survey?

Answer 12

• precise behaviours (log data is better for this, but also not 100% reliable)
• underlying motivations
• usability evaluations (usability studies better, which is observation based)

Question 13

Drawbacks of a survey?

Answer 13

• we need to consider our design and confounding factors e.g including multiple dependent variables in a question
• low completion rates (repeatable questions , poor design,poor questions)
• noisy data from bad question design (e.g. vague / ambigous / recall)
• biased questions