Developing a questionnaire

Question 1

Define RESEARCH

Answer 1

A study which is trying to generate new knowledge about patients/conditions/healthcare services that will be generally true across a wider population than it is actually studying

Question 2

What is service evaluation?

AKA Service improvement/quality improvement

Answer 2

Assessing a part of a local healthcare service with the aim to improve the service based on the evidence that is collected

• Results will only feedback into this particular service and will not apply to patients/service nationwide
• Methods vary in service evaluation

Question 3

What is a clinical audit?

Answer 3

A continuous process o quality improvement in which you IDENTIFY what it is you want to achieve, adopt a METHOD to collect information about the service, ANALYSE whether the service is doing as it should, make a CHANGE, MONITOR its effect.

Question 4

State three methods used to find out information for a clinical audit?

What should you consider when picking a method?

Answer 4

• Patient notes/records review
• Survey of staff (paper/online)
• Survey of patients (paper/online)
• Telephone interview
• Face to face interviews
• Focus groups

Logistics, anonymity, consent, cost, data protection

Question 5

Give three issues around patients’ notes reviews

Answer 5

• No patient identifiers can leave clinical environment so anonymise notes immediately if possible
• Limited number of types of information available
• Different types of data e.g. categorical (Y/N) or continuous (BP) which will determine type of analysis

Question 6

How can decide which categories to use in questions?

Answer 6

Stick to standardized categories outlined by Office of National Statistics

Question 7

What factors determine the format of questions?

Answer 7

• What we know already
• What type(s) of data we want to generate
• What type of analysis. we want to do

Question 8

What type of questions might you include in a questionnaire?

Answer 8

Closed questions with Y/N or categorical answers

Closed statements answered on a scale (e.g. strongly agree to disagree)

Open ended questions

Multiple choice with one or more options

Question 9

How can you make you questions as clear as possible?

Answer 9

• Use simple direct language
• Be specific
• Break down big ideas into multiple questions
• Avoid leading questions

Question 10

What are Likert scales and when might you use them>

Answer 10

A questionnaire method in which respondent specify level of agreement on a symmetrical scale for a series of statements.

Used for knowledge, belies, attitudes, symptoms, traits

Question 11

What are the 5 facts/assumptions with Likert scales?

Answer 11

1. Uses a bipolar scaling method, measuring negative or positive responses to a question
2. Assumes the interval between each response is equal (interval level data)
3. Each response gets a number 1-5
4. Usually multiple statements, related to each others, are posed
5. Scores are summed across statement sto form a scale- all items have equal weight

Question 12

What should you consider when writing statements for Likert scales?

Answer 12

• Is the question/statement loaded with social desirability? (therefore increasing bias as respondent provides a more socially acceptable answer)
• Does the question require knowledge the respondent might not have ?
• Is the statement based on two premises?

Question 13

Why is it important to include many items into your questionnaire?

Answer 13

• Youre unlikely to identify a single item which represents the whole issue being measures
• A spread of questions represents every aspect of the concept
• Redcues error of random responses or misunderstandings
• Can design questionnaire to reduce skew

Question 14

How should you approach analysis of answers?

Answer 14

• Consider your study questions
• Plan analyses to address study questions as part of initial study design
• Can convert questionnaire responses to numbers if quantitative analysis.

Question 15

If you convert questionnaire answers into quantitative responses what must you consider?

Answer 15

• Will you include binary categories? E.g. Ethnicity 1= white, all other responses =0
• You must record how you will score each questionnaire item BEFOREHAND

Question 16

How can you analyse likert scare responses?

Answer 16

Add up scores across similar questions to make a scale (total score)

Report mean total score and sd

Question 17

How could you analyse open ended questions?

Answer 17

Qualitative thematic or framework analysis

Question 18

What does comparing groups’ responses tell us?

e.g. do men answer differently to women?

Answer 18

• Compares frequency of occurence
• For yes/no questions: Chi square test to see if there is a significant difference in frequency
• For likert questions: T- test of mean difference

Question 19

What 4 parameters must be considered in recruiting a representative sample of service users?

Answer 19

HOW will you recruit?

WHAT will be the sampling technique?

• random/stratified sampling?
• convenience/self identified/consecutive?

HOW will you ensure a good spread of participants? To reflect the population

HOW will you ensure a low dropout? Reporting non responders, considering factors for dropping out

Question 20

What 4 parameters of data management must be considered?

Answer 20

Data recording
- How will you deal with missing data?
Data entry
- Quality control
Data storage
- Excel?
Data security
- Password protected database, anonymisation, locked paper copies