data collection part 3 of 4

Question 1

what are the 2 different types of studies which need to be considered when deciding the type of data collection

Answer 1

prospective

retrospective

Question 2

define prospective study

Answer 2

record variables over the the study period with the outcome measure subsequently.

Question 3

define retrospective study

Answer 3

measure the outcome and then look bad to measure the exposure and other variables.

Question 4

An example of a prospective study

Answer 4

cohort

Question 5

An example of retrospective study

Answer 5

case control study

cross sectional study

Question 6

what are the pros of a prospective study

Answer 6

• fewer sources of bias (no recall bias)
• less chance of confounding
• some data can only be measure prospectively

Question 7

what are the cons of a prospective study

Answer 7

• time and resource intensive
• tendency to collect data on more variables than you need or can use in your analyses.
• subjects can drop out (bias)
• unfeasible for rare outcomes

Question 8

what are the pros of a retrospective study

Answer 8

• less time and resource intensive

- allows oversampling of rare outcomes

Question 9

what are the cons of retrospective study

Answer 9

• more suitable to bias (recall + recruitment)
• some variable cannot be measure directly (e.g. BMI, BP, pre outcome/ pre treatment)
• if data are form records little control over these

Question 10

what is the purpose of a research diary

Answer 10

• ensure timely record of what you do

- observations on data collection processes influencing data quality and completeness.

Question 11

what id the purpose of a data collection pro forma

Answer 11

-variables are measures and recorded CONSITENTLTY.

Question 12

What is a commonly used pro forma

Answer 12

questionnaire

Question 13

what are 3 common pit falls in questionnaires

Answer 13

• ambitigious questions (would you normally feel uncomfortable about your diabetes.
• leading questions ( to what ethnic group do you belong
• multiple questions (have you found diabetes to be limiting longstanding illness)

Question 14

what are 2 types of questions used in a questionnaire

Answer 14

• open ended (respondent determined)- need coding prior to analysis
• closed ended (part respondent/part researcher determined)

Question 15

what does piloting proformas mean

Answer 15

Ask non-participants/researchers to try using draft versions of the questionnaires/proformas – are they:

• ambiguous?
• leading?
• multi-factored?
• undermined by missing answer categories?
• ‘face valid’ (i.e. do they make sense)?

Try ‘coding’ and/or capturing data from a sample of completed questionnaires/proformas – are there:

• missing answers?
• inconsistent answers?
• implausible answers?