Discussion
Weaknesses
One weakness of my study was
response bias. Out of 80 surveys passed out, about 70 were
returned. More boys returned the survey than girls, so in order
to make the number of girls and boys included in the study equal, some
surveys had to be discarded. Also, some people failed to answer
the questions with numerical answers, which was required and so, these
also could not be included in the study. Also, some students
failed to understand that there are only twelve inches in a foot and
recorded their height with inches over twelve, so these also had to be
discarded.
Another issue was selection bias. After obtaining the sample,
when the schedules were looked up, there were two male students whose
schedules were not included in the schedule box, so in both cases, I
took the student whose name was directly after the students I was
looking for in alphabetical order.
Genetics is also a major factor in height and cannot be very easily controlled in a statistical study naturally leading to this study being weak.
Extrapolation
I think that despite the errors in the
study, the overall results should not have been affected very much and
so it would be safe to extrapolate to all high schools in the area
since high school students are fully grown and so it could be expected
that there will be little to no relationship between students’ current
hours of sleep and height. Also, while there was some response
bias, it was minimal, and only included a few people so effects on the
results should have been very few.
Suggestions
If someone should do a study like this, I
think it would be better to choose a younger population who is still
growing in height since that is when a relationship between hours of
sleep and height would be visible if there is one. Also, since
high school students are no longer growing, their current hours of
sleep would not be affecting their height, but rather their sleep
patterns during their period of growth which may have been very
different from their current ones.