Quest Ver2

Questionnaires: Making questions answer questions

Sir Francis Galton, F.R.S. is known as the father of the questionnaire. He was also a half-cousin of Charles Darwin, an English Victorian polymath, anthropologist, eugenicist, tropical explorer, geographer, inventor, meteorologist, proto-geneticist, psychometrician, and statistician.

How to analyze

WhenWhat Kind

One Best

Source

The British Medical Association, considered one of the most authoritative research organizations, publishes a peer-reviewed journal called BMJ (formerly British Medical Journal).

In 2004 BMJ published a series entitled "Hands-on guide to questionnaire research“.

When to Use Research

Questionnaires

Best Method of Data Collection for Chosen

Subject?• Pre-requisites for using questionnaires:

– Well defined concept of what information you actually want to collect

– A good grasp of the topical territory

• When a questionnaire is not the best method

– http://www.bmj.com/cgi/content/full/328/7451/1312/DC1#a

Standardization• Meaning of “Standardized”

questionnaire:– Identical questions– Identical format– Identical response recording process

• Standardization creates reliability– Do not use a non-standardized

questionnaire unless you want unreliable results

Can You Use Existing Designs?

Published, authenticated questionnaires are invaluable:– Already standardized.

– Allow comparison of your findings with other studies that used the tool.

– Saves time of having to design and test a new questionnaire.

http://www.bmj.com/cgi/content/full/328/7451/1312?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&

fulltext=Boynton+PM%2C+Wood+GW%2C+Greenhalgh+T.+Hands-on+guide+to+questionnaire+research%3A+reaching+beyond+the+white+middle+classes&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT

How Reliable is the Design?

• Participants answers may not reflect reality– Example: Self-completed questionnaires on

dessert intake only measure what participants say they have ingested not what has actually been consumed.

• Must account for original context of questionnaire– Time period– Culture– Country, etc.

Types of questionsOpen questions: answers not preset; allow participants to create own responses

Closed: give participants options e.g. check boxes, mark a scale, select an answer from list, etc.

Pros and ConsOpen questions:• pros: potentially unlimited range of answers• cons: more difficult to translate into data--

takes a lot more time and effort to do

Closed questions:• pros: easy to convert answers to empirical

data• cons: can be frustrating to participants,

potential range of answers limited by researchers

Types of closed

questions1.Statements

with tick box categories □ yes □ no

Please check the box that best agrees with your answer

2. Numerical Rating Scales

4 3 2 1 Please indicate

how convenient you find our new hours

3. Visual analogue

scales

On the line below please indicate with an X how you have reacted to your new medication

You have just received a lesson in healthy eating. Please look at the faces above and pick the one that best symbolizes how you feel about this advice.

4. Symbols

Before beginning study:Budget both time and moneyAgree on program and coding

system

“It’s better to have fewer good quality responses than a high number that are inaccurate or incomplete.”

Analysis

Descriptive

Statistics ☛ Averages, Ranges, Ratios

X # said X

☛ Complex comparisons

Inferential Statistics

Spearman's rank-order correlation.

Pearson's coefficient of correlation

Chi-square distribution

Simple interrelationships

Analysis of Variance (ANOVA) multiple variables can be tested at the same time : 3 or more variants can be analyzed simultaneously with a pivot table

Multivariant analysis

♀ responded X

♂ responded X

Methods section1. Details of exclusions

2. Data on refusal/unsuitable3. State and justify statistical analysis used

1. Ensure results are hypothesis driven2. Report non-significant results

Results section

1. Suggest what findings mean2. Acknowledge limitations3. Discuss the questions participants

refused to answer or answered in an unpredictable manner

Discussion section

The End ?

??

? ?

??

???

?

??

?? ? ?

?