Limitations
Like all studies, this study has limitations.
For one, quasi-experimental designs, like the one
used in this study, are susceptible to internal validity
threats.37 To contend with this concern, selection
bias was minimized through the use of a balanced
design. Attempts were made to ensure the two
groups had proportional representation with regards
to extraneous variables.37 Pair matching was used to
control the effects of GDM classification and prepregnancy BMI. Comparison of personal characteristics
of both the experimental and the control groups,
before initiation of the self-regulation program was
executed, resulted in no significant differences.
Attrition bias was addressed and the two control
group subjects, who left the study after the fourth
week of the program, were found to not be
significantly different from those who remained.
Therefore, the self-regulation program, rather than
other factors, affected diet control, exercise behaviors
and two-hour postprandial blood glucose of the
experimental group subjects. However, the findings
of this study can be generalized only to pregnant
women with GDM who have characteristics similar
to those used in the study.
Another study limitation is the fact the data
relied on self-reports of food consumption, exercise
and two-hour postprandial blood glucose levels.
Thus, one has to assume the subjects’ recordings
were accurate. Self-reports are vulnerable to the risk
of reporting biases, but tend to yield data of
considerable richness that is useful in gaining an
understanding of the phenomena under examination.
Finally, subjects were obtained from only one
geographic area (Bangkok) in Thailand which was a
major metropolitan area. Thus, the findings are
applicable only to pregnant women with GDM who
are from Bangkok and not to women from rural areas
of Thailand.