Sourcing Data

Question 1

Define ‘data sourcing’.

Answer 1

How one goes out to gather data.

Question 2

Which two ways of data sourcing are there?

Answer 2

Active and Passive

Question 3

What is active data sourcing?

Answer 3

These are ways of acquiring records (or data) actively through fieldwork, surveys, interviews etc.

Question 4

What is passive data sourcing?

Answer 4

These are ways of acquiring records (or data) that are made available from some “official” (or “recognised”)
sources (e.g., NGOs, agencies, companies, government, educational institutions etc.)

Question 5

What is Open Data?

Answer 5

Official/recognised sources can make their data freely available to everyone to use & republish, repurpose it as they wish without any patent, legal or copyright restrictions.

Question 6

What type of data is Open Data?

Answer 6

Secondary Data

Question 7

What type of data is active data sourcing?

Answer 7

Primary Data

Question 8

Define primary data.

Answer 8

This type of data source refers to the first hand data gathered by the user, researcher or enumerator through fieldwork, interviews, questionnaire surveys etc.,

Question 9

Define primary data.

Answer 9

First hand data gathered by the user or researcher through fieldwork, interviews, questionnaire surveys etc.,

Question 10

Define secondary data.

Answer 10

Data that has already been collected through a primary source and made readily available for other researcher(s) to use for their study or investigation.

Question 11

What is internal secondary data?

Answer 11

If you are working in collaborating with an organisation that has some relevant data of interest and the organisation has collected the data and can provide
access to such users working within it or collaborators.

Question 12

Give an example of internal secondary data.

Answer 12

Data scientists working within the UK Metropolitan Police service seeking to quantify the burden of various crime outcomes e.g. burglary, sexual assault etc.

Question 13

What is external secondary data?

Answer 13

• Open source websites (freely accessible)
• Paying for the data (requires a license)
• Online data service which is free but requires users to
  register etc.

Question 14

List the advantages of primary data sources.

Answer 14

• Data collected is always up-to date
• Relevant and specific to user’s research aims and
  objectives
• High-level, and greater of understanding about the
  nature and content of the dataset
• High-level of accuracy as along as you do whatever in your power to minimise all kinds of systematic errors in the data collection process

Question 15

List the disadvantages of primary data sources.

Answer 15

• Depending on the study design – data collection is a very time consuming and expensive process
• If you are collecting personal and sensitive data, you must DEFINITELY apply for ethnical approval before going out to get your data
• You will have to clean, manage and maintain your own data
• Possibility to falsify his/her data since one has his/her autonomy over the data

Question 16

List the advantages of secondary data sources.

Answer 16

• Ease of access and low cost, or even free if the data is from an open source platform.
• Time-saving, especially if the data has already been processed and cleaned
• Unlike primary data; secondary data are often
  collected routinely hence allowing for longitudinal analysis
• Often secondary data are combined with different
  sources making it have variety of variables.

Question 17

List the disadvantages of secondary data sources.

Answer 17

• No control over the quality
• The secondary data might not be specific to your needs
• The data can be biased in favour of the one whose gathered it.
• You are not the owner and will never full understanding of the its nature & how it was collected

Question 18

Define ‘research design’.

Answer 18

The investigator’s plan for tackling a research question through collection of data, analysis and interpretation of such data, and finally a thorough discussion of that said data.

Question 19

List the two types of research designs.

Answer 19

Qualitative and quantitative.

Question 20

Describe ecological study design.

Answer 20

Unit of observation are at group- or aggregated level

Question 21

Describe a cross-sectional study design.

Answer 21

Data is collected on a single snap in time, and its
useful for generating further hypothesis.

Question 22

Describe a case control study design.

Answer 22

Observe the effect of an independent variable(s) on an
outcome that has already occurred. The time frame is
a retrospective analysis.

Question 23

Describe a cohort study.

Answer 23

Observe the effect of an independent variable(s) on an
outcome that has not happened yet. The time frame is
a prospective analysis.

Question 24

Describe a randomised control trial (RCT).

Answer 24

Observe the effect in an actual test group(s) (e.g.,
intervention, effectiveness of a teaching programme,
clinical trials for a vaccine)

Question 25

Whats the cost effectiveness of an ecological study design?

Answer 25

Cheapest as it relies on routinely collected data most of the time.

Question 26

Whats the cost effectiveness of a cross-sectional study design?

Answer 26

Less expensive as its conducted at a single time point and requires less resources.

Question 27

Whats the cost effectiveness of a case-control study design?

Answer 27

Quite expensive to interview participants to provide past experiences.

Question 28

Whats the cost effectiveness of a longitudinal study design?

Answer 28

Most expensive as it requires two or more follow up of subjects enrolled in the study so more resource and time are required.

Question 29

List a common bias involved in an ecological study design.

Answer 29

Ecological fallacy

Question 30

List a common bias involved in a cross-sectional study design.

Answer 30

Results are only representative at time of study.

Question 31

List a common bias involved in a case-control study design.

Answer 31

Recall bias.

Question 32

List a common bias involved in a longitudinal study design.

Answer 32

People dropping out of study can introduce lost-to-follow-up bias.