2. It involves:
selection
training
supervision
Validation
Evaluation
Of the persons involved in the collection of data
3. SELECTION
Involves careful selection of people who will
conduct the interviews or the respondents from
where data is going to be collected
Type of data to be collected will influence the
type of persons to be recruited e.g. qualitative
data that involves probing questions would
require experienced interviewers
4. TRAINING
Involves training personnel that will go out to
collect data
Training helps to achieve uniformity in data
collection
Training covers aspects such as the approach,
administering of instruments, probing,
recording
5. SUPERVISION
Interviewers need to be supervised to ensure
that they are proceeding on schedule
There is need to keep control of day-to-day
activities
Unethical practices might occur e.g. when field
workers complete questionnaires themselves.
6. VALIDATION
Quality of data collect is critical
therefore there is need for
supervisor to check accurately for
mistakes that might occur during
data gathering and recording.
7. EVALUATION
All data gathered have to be evaluated to
check for consistency.
How many people completed the
questionnaire, quality of the interviews
Two errors might occur:
Intentional errors- falsehood and non-
response
Unintentional errors- misunderstanding,
fatigue, attention loss
8. The process of checking the quality of data
gathered during fieldwork, thus converting it
into useful format for further analysis
Involves:
Validating
Editing
Coding
Data entry
Data cleaning
9. VALIDATING
An analysis to check whether proper
procedures were followed in the collection of
data using different instruments
Factors to look for include:
Fraud
Screening
Procedure
Completeness
Courtesy
10. EDITING
Checking of mistakes emanating either from
interviewer or respondent
Editing can start when the supervisor visits the
fieldwork when data is being collected and
after completion of questionnaires for
omissions and inaccuracies
Further editing might be done at the central
office of the research firm
11. CODING
Process of assigning a code, symbol, or a
number to each possible answer.
It is easier for closed ended questions where
pre-coding can be conducted
Closed ended questions would require coming
up with themes as same questions can be
responded to differently but with the same
meaning
12. CODING-cont’
Missing data is usually caused by:
The question did not apply to the respondent
The respondent refused to answer the question
The respondent did not know the answer to
the question
The respondent or interviewer forgot to record
the answer
13. DATA ENTRY
Involves the direct input of coded data into
software packages that will allow the
researcher to analyse and transform the raw
data
Software packages include SPSS for
quantitative data and NVIVO 7 for qualitative
data
14. DATA CLEANING
It is the process of checking data
before starting data analysis
especially if data was entered
manually in a computer
15. DATA CLEANING- cont’
Done to detect obvious errors such as:
Does the number of questionnaires match the
number of respondents
Calculation of minimum and maximum variables to
check whether they are outside the expected range
Is there consistency in the number of respondents
who provided answers
Comparison of recorded answers with the values in
the data matrix for a few selected questionnaires
16. Options for large percentages of missing
answers:
Casewise deletion- discard any cases with
missing responses
Pairwise deletion- use all the cases with
complete responses for specific calculation
Rules of thumb- establish your own rules as
to when to include or discard individual
questionnaires