Microsoft Office Tutorials and References
In Depth Information
3.2 Hypothesis Testing
One of the important activities of research scientists is that they attempt to
‘‘check’’ their assumptions about the world by testing these assumptions in the
form of hypotheses.
A typical hypothesis is in the form: ‘‘If x, then y.’’
Some examples would be:
1. ‘‘If we use this new method fertilizing the soil, the corn yield of the plot will
increase by 3 %’’.
2. ‘‘If we use this new method of teaching science to ninth graders, then our
science achievement scores will go up by 5 %’’.
3. ‘‘If we change the format for teaching Introductory Biology to our
undergraduates, then their ﬁnal exam scores will increase by 8 %’’.
A hypothesis, then, to a research scientist is a ‘‘guess’’ about what we think is
true in the real world. We can test these guesses using statistical formulas to see if
our predictions come true in the real world.
So, in order to perform these statistical tests, we must ﬁrst state our hypotheses
so that we can test our results against our hypotheses to see if our hypotheses
match reality.
So, how do we generate hypotheses in science research?
3.2.1 Hypotheses Always Refer to the Population of People,
Plants, or Animals that you are Studying
The ﬁrst step is to understand that our hypotheses always refer to the population of
people, plants, or animals under study.
For example, if we are interested in studying a species of noxious weed found
along highways of southern South Dakota, we would select various sections of
highways and estimate the number of weeds found in these sections, these sections
would be used as our sample. This sample would be used in generalizing our
ﬁndings for all of the highways in southern South Dakota.
All of the highways in southern South Dakota would be the population that we
are interested in studying, while the particular sections of highways in our study
are called the sample from this population.
Since our sample sizes typically contain only a portion of the highways, we are
interested in the results of our sample only insofar as the results of our sample can
be ‘‘generalized’’ to the population in which we are really interested.
Search JabSto ::

Custom Search