Alswh assesses womens physical and mental health

• Turn in your short write up as a .pdf file by 11:59 pm AEST on 17 June. Email the report directly to me:

SURVEYS

SURVEYS

Describe what is happening. Example:
– Identify characteristics of target markets (who, what, where, when, why).

4. We would appreciate your reaction to Buskers by the Creek so far. How would you rate the following?

ERRORS IN SURVEY RESEARCH

Suppose I wanted to know the average height of Bond students.

Let’s start by discussing sources of error in survey research.

IMPORTANT!

VARIATION DUE TO SAMPLE SELECTED (RANDOM SAMPLING ERROR)
How tall are Bond University students?

ERRORS IN SURVEY RESEARCH

‘reasonably big’ and properly selected)

(We will return to samples later…)

IMPORTANT!

Customer feedback tends to be bi-modal – opinions tend to be extreme. Why is that?

• No contact: A person who cannot be reached – phone turned-off, they moved...

• Refusal: Person unwilling to participate in
research (participation rate may be 1-2% if spammed on the internet; 30% in person).

Try to keep a count of ‘Non-respondents’ (number of refusals and non-contacts) and report that percentage.

Have you had any drinks

today, sir?

Source: National
Geographic, June
2017

• Acquiescence bias: some individuals tend to agree with all questions or concur with a particular
position.

• The Asch experiment:

A "yes" vote supported legalizing recreational marijuana for adults over the age of 21.

A "no" vote opposed this proposal to legalize recreational marijuana.

RESPONSE BIAS (CONTINUED)

• Extremity bias: some individuals tend to use extremes when responding to questions, like checking all 7s (some tend to use the middle option!).

RESPONSE BIAS (CONTINUED)
Social desirability bias is caused by respondents’desire to gain prestige or appear in a different social role.

“Oh yes, I go to the opera all the time.”

source: INDEPENDENT, 20/2/16

Collectively, couples presumably do 30 hours of housework per week?! BTW: What IS housework? Dropping children at school? Planting vegetables? Exploratory qualitative research could be used to provide insight – I suspect people (men/women) have very different views of what is housework!

(the second “systematic error” bucket)

• Data processing error: An error that results from incorrect data entry.

• Interviewer error: Mistakes made by
interviewers who fail to correctly record survey responses.

• Interviewer cheating: Filling in fake answers or falsifying questionnaires. How could we check for this?

QUESTIONS
• A structured question (closed –ended) imposes a limit on the number of allowable responses, e.g., 1-7 scale, yes/no, Apple/Samsung/Other.

• An unstructured question (open-ended) does not restrict the respondents’ answers.

TEMPORAL CLASSIFICATION• Cross-sectional study: Collects data at a single moment in time, as you will do.

– Consumer panel: A study of the same sample of individuals or households to record their attitudes, behaviour or purchasing habits over time.

• (see next slide)

And usually higher response rates…

SHOPPING MALL INTERCEPTS• Mall intercept interviews are conducted in a shopping mall.

• Computerised voice-activated: Done without human interviewers.

– works best with very short, simple questionnaires.

SELF-ADMINISTERED QUESTIONNAIRES

• Survey in which the respondent takes the responsibility for reading and answering the questions.

RESPONSE RATE

How to increase response rates for mail surveys– Stamped return envelope
– Attractive questionnaires
– Personalized cover letter
– Monetary incentives or premiums
– Meaningful topic/Interesting questions
– Follow-up reminders
– Advance notification
– Survey sponsorship

SELECTING THE APPROPRIATE SURVEY METHOD

• Decreases likelihood of respondents
misunderstanding a particular question or misinterpreting instructions.

How often do you read the newspaper?

– Respondents have a right to refuse to participate or to answer specific questions.

– Researchers need to protect the
anonymity/confidentiality of the participants.

A sample explanatory statement

• Pretesting prior to actual administration allows for correction and improvement of survey.

• Be ethical! Ethicality must be preserved to ensure research and researcher’s integrity.