Adasd

DESCRIPTIVE
RESEARCH
Chief characteristics
• Due to the nature of the classification variable (i.e., a subject variable), extraneous variables affecting the classification variable cannot be controlled;
• However, extraneous variables affecting the measurement of the criterion variable can usually be controlled, although this may not be desirable for very high external validity;

Slide 1

• No manipulation of independent variable;
• Subject groups are not the result of random assignment, through they may be designated through random selection;

Slide 2

• Aims to examine:
1.
2.
3.
4.

differences between groups relationships among variables trends (changes over time) status or level of a variable

• Relative to experimental research, internal validity is low, while external validity is high;
• Research hypothesis may be of less importance - descriptive research often aims to generate hypotheses which may then be experimentally tested;

Slide 3

• Data may be quantitative or qualitative and may not be amenable to statistical analysis. Descriptive research "attempts to provide an accurate description of a particular situation ... to identify the variables that exist in a situation/to describe the relationships existing between
(among) these variables." from Christensen, Experimental Methodology

Slide 4

VARIABLES
Independent variable
In experimental research, the researcher manipulates the I.V., i.e., applies or withholds it; varies its level; etc.

Dependent variable
The
variable measured experimenter in order to detect

by

the

Classification variable
In descriptive research, subjects are classified according to the classification variable, e.g., by sex, age, disorder, etiology, etc.

Slide 5

Criterion variable
The classified subjects are compared according to some criterion, e.g., rate of learning, hearing threshold, vocabulary age, etc.
In descriptive research, the classification variable is analogous to the I.V. of experimental research, and the criterion variable to the D.V. Some researchers prefer to use I.V. and D.V. instead even in descriptive research.

Slide 6

Predictor and predicted variables
In correlational studies, analogous to
I.V./classification variable.
In such studies, the predictor variable is used to predict the behavior of the D.V./criterion variable, which is designated the predicted variable. E.g., an auditory discrimination measure may serve as a predictor variable for an articulation measure, the predicted variable. Slide 7

DESCRIPTIVE
RESEARCH
Principal types
Comparative
Aims to measure the behavior of a criterion variable at a single point in time in order to draw conclusions about similarities or differences between the classification variables (i.e., between the populations). Correlational
Aims to measure the degree to which a predictor and predicted variable are associated/related, i.e., the degree to which changes in the predictor correspond to changes in the predicted.
Slide 8

Developmental
Aims to measure changes over time
(trends) in the criterion variable (e.g., due to ageing, maturation, recovery, learning, etc.). Longitudinal
Studies one group of subjects over a prolonged period of time.
Cross-sectional
Studies two or more groups of subjects, each of which is at a different, progressive level (i.e., age, stage of recovery or learning, etc.).

Slide 9

Descriptive
Aims to ascertain the level or status of a criterion variable at a specific point, without comparison of groups, or of stages of development (= description only).

Slide 10

SYSTEMATIC OBSERVATION
Steps
1. Choose the criterion variable (i.e., a natural behavior).
a. specific
b. unambiguous
2. Select setting
a. where criterion variable can be expected b. where observer can remain unobtrusive Slide 11

3. Decide on mode of recording
a. live or delayed
b. coding - recording sheets
c. instrumental
4.
Decide on observation sampling strategy a. time-point
b. time-interval
5. Train all observers

Slide 12

NATURALISTIC OBSERVATION
"Fly on the wall"

3 dimensions
Natural behavior (criterion variable)
Natural setting
Natural "treatment" (classification variable) Slide 13

PARTICIPANT OBSERVATION
Common stages (from Kidder)
1. Selection of venue or situation
2. Gaining access to setting
3. Taking role in setting
4. Making notes
5. Analyzing notes

Slide 14

Steps in construction of questionnaire or structured interview 1.
2.
3.
4.

Decide type of information sought
Decide on type of questionning
Prepare 1st draft
Seek suggestions/criticism from others; revise if necessary
5. Test questionnaire/interview on small
"pilot" group
6. Edit and revise based on results or suggestions from pilot respondents.
7. Prepare final version - decide on procedures for distribution and for use.

Slide 15

Brief checklist about questionnaire or structured interview questions
Content
• Necessary/useful?
• Is one question enough?

• Is question adequate to cover area of interest? • Does it assume too much information on part of respondents?
• Is the content too specific to a particular category of respondent?
• Is the content too specific or too general for aims?
• Is the content biased?
• Will the respondents be willing to answer? Slide 16

Wording
• Unclear phraseology?
• Jargon?
• Questions too long/complex?
• Unstated assumptions or unseen implications? • Biased or emotionally loaded terms?
Stereotypes?
• Objectionable wording?

Slide 17