SWP32RES Research for Social Work Practice B

1
SWP32RES Research for Social Work Practice B

• Lecture 6
• Data Collection Methods

2
Data collected in human services research will be
collected in two broad forms either

• 1) Quantitative Symbols (eg. numbers, ,
  priority rankings, statistical test outcomes)
• 2) Qualitative Symbols (ie. words).

3
There are 4 broad ways in which data can be
gathered in human services research

• 1 Interviewing
• 2 Responding to written questions
• 3 Observing
• 4 Examining secondary data

4
When do you use a questionnaire and when do you
use an interview?

• A questionnaire should be chosen when
• - large numbers are involved
• - facts rather opinions are sought
• - data in a standard format is required
• - wide geographic coverage is required
• - time for the respondent to reflect could be
  useful

5
- long lists of statements need to be
read - researcher time is at a premium
6
An interview should be chosen when

• - smaller numbers are involved
• - the enquiry is exploratory
• - attitudes are sought
• - complicated reasons for actions are sought

7

• - sensitive areas are being explored
• - complex situations exist
• - non-verbal responses could be significant
• - flexibility is required (Fuller Petch
  (1995, p.56)

8
Interviews provide a researcher with a number of
advantages. These are

• 1. Opportunity to probe - opportunity to expand
  on responses more fully thereby gain more
  in-depth information. Helps to understand
  individual's attitudes and perhaps its origins.
• 2. High Completion rate - if someone agrees to be
  interviewed they usually complete the actual
  interview.

9

• 3. Access is provided to additional information
  such as nonverbal behaviour, the respondents
  attitude to the interview, home environment (if
  conducted in their home), and their interaction
  with significant others (based on Gochros, 1988,
  pp.267-9).

10
But there are also disadvantages associated with
interviews

• 1. Influence of researcher - very presence of
  interviewer will do this (the so-called
  expectancy effect).
• 2. Potential for recording errors - data can
  be forgotten, distorted or misinterpreted.

11

• 3. Expense - for training of interviewers and
  travel to respondents (based on Gochros, 1988,
  pp.26-73).

12
There are three choices on the structure of an
interview schedule

• 1. Structured
• 2. Semi-structured
• 3. Unstructured (in-depth)

13
3 Observing

• Observation revolves around two variables the
  nature of the observer role and whether it is
  structured or unstructured.
• Observer roles vary along two dimensions
  concealment and degree of participation.
• Hence roles can have the following combinations

14

• - concealed, nonparticipating (complete
  observer)
• - concealed, participating
• - not concealed, nonparticipating
• - not concealed, minimal participation
• - not concealed, participating (participant
  observer)

15

• But there are also disadvantages and problems
  associated with observation.
• Kellehear (1993, p.128) provides a comprehensive
  list of these

16

• 1. Observer bias resulting from selectivity
  (the unusual attracts whilst the familiar is
  taken for granted )
• 2. Observer bias resulting from other sources
  like fatigue and boredom
• 3. Inadequate or biased recording

17

• 4. Generation of large volumes of data
• 5. Complex or busy patterns which can confuse
• 6. Emic/etic (insider/outsider) problems if
  observation used alone
• 7. Limited usefulness in public places

18

• 8. Population instability
• - difficulty of keeping tabs on chosen
  subjects
• - different environments contain different
  people at different times
• 9. Recall problems if recording not immediate

19
4 Examining Secondary Data

• There are particular advantages to secondary data
  analysis. These are
• 1. Provides a potentially comprehensive source
  of data which is not necessarily restricted to
  secondary sources alone.

20

• 2. Data are often unique
• 3. Often supplies good sources of longitudinal
  data
• 4. Data are reliable ie. can easily be
  rechecked and verified by others (These
  points derived from Daly, Kellehear Gliksman,
  1997, p.126)

21

• 5. Financial costs can be minimal.
• 6. Nonintrusive.
• 7. Relatively hassle-free in gaining access
  and permission.

22
There are methodological problems related to this
form of analysis including

• 1. Generalisability may be limited
• 2. It artificially categorises ambiguous and
  changing processes
• 3. Difficult at times to establish
  authenticity
• 4. Degree to which original measurements are
  accurate - can the data be trusted? Is it
  biased?

23

• 5. Data that one can secure in a study are
  limited to those recorded in the original
  study report or other document. Sources may
  be incomplete.
• 6. It can decontextualise behaviour and
  meaning. (Based Daly, Kellehear
  Gliksman, 1997, pp.130-1)