Professional Documents
Culture Documents
Disadvantages
since setting is completely natural, with all variables present, cannot identify cause
Advantages
can acquire a lot of information through description
useful for identifying variables & hypothetical constructs > which can be further
investigated through other means (kind of like exploratory surgery, or the USS
Enterprise on an away mission)
1) Field Surveys
subjects complete a questionnaire or interview in a natural setting
to increase return rate: include a cover letter, ensure survey has face validity,
use only brief questions with simple instructions, use a small reward, send a
reminder letter
Telephone: structured & closed ended questions (no free associations please)
usually higher response rate
subjects who don’t participate may be too busy or too wary
make sure intro is straightforward & friendly
cluster sampling: certain groups are randomly sampled & all subjects in them
are observed
convenience sampling: study subjects who are convenient
quota sampling: using convenience sampling, with the restriction that the
sample has the same % of each subgroup as that found in the
pop
snowball sampling: “and they told two friends, and so on, and so on...”
Potential biases:
time of telephone call
volunteer bias (people with strong opinions, or bored, are most likely to respond)
the silent majority?
predictive validity? How you feel today may not be how you feel tomorrow
use descriptive stats to check for significant correlations between responses from different
questions (see if significant using inferential stats)
2) Observational Studies
descriptions based only on unobtrusive (covert) observations of the subjects
3 general techniques
naturalistic observation
unstructured, unsystematic observation of a wide range of behaviors or situations
participant observation
the investigator “infiltrates” the group being observed (like a spy)
takes part in group activities but true identity and purpose are concealed
Advantages
100 % natural > no subject reactivity or demand characteristics
Disadvantages
1. experimenter expectations (esp in participant observation)
2. lack of informed consent
3. usually have to rely on convenience samples
4. mostly qualitative data: verbal descriptions based on visual observations
may lack precision & accuracy (observer bias) & not very sensitive
Purpose
to identify variable which might be important
indirect tests of models or predictions
the “data” in descriptive research designs is frequently based on observations of actual behavior or
the responses to questions ... but there other types/sources of data ...
archival research:
written records
allows access to otherwise unobservable behaviors
can verify subjects’ self-reports
case studies
detailed observations of one situation or one person
case study (longitudinal, prospective): subject to time-related confounds
case history (archival, retrospective): poor reliability & validity
Program Evaluation
1. needs assessment
identify what’s needed & whether the program will be used
2. program planning
design the program - methods
3. program monitoring
make sure program is implemented and clients are using it
4. outcome evaluation
is the program having the intended effect?
data for each phase collected through descriptive research methods