The most important part of any research project is the
planning process. This statement is as true for data analysis
as for any of the other steps in the research process. The
development of your statistical analysis plan should not be
delayed until after you have your data in hand. Rather, the
investigator should select the statistics to describe the sample
and to analyze the data for each research question or
hypothesis before initiating the study. Most grant applications
will require this information, but these decisions
should be made regardless of application for funding.
Investigators should plan to first describe their sample.
They should identify the important demographic characteristics
of the sample, such as sex, age, and race. These variables
will be the same for most studies. Other sample
characteristics, such as diagnosis, weight, height, Glasgow
Coma Scale, and so forth, also may be important to provide.
Descriptive statistics will describe these variables.
Next, investigators should plan the analyses for each
research question/hypothesis. A table may be useful for this
activity. In the first column should be the research question/hypothesis;
in the second, all relevant variables (and
timing information if needed); and in the third, the statistical
test to be used. This process helps investigators ensure
that they are collecting all needed data, at the right time, to
answer their question. After all study data have been collected
is not the time to discover that an important piece of
data has been missed.
Investigators uncomfortable with statistical analysis should
consult a statistician early in the planning phase. A statistician
will help them determine what statistical analyses are most
appropriate for answering the research questions/hypotheses,
taking into consideration the types of data to be collected.
Although statisticians may seem intimidating, investigators
should consider them an important member of the research
team and avail themselves of their expertise.
To help diminish the stress of a statistical consultation,
investigators should prepare a list of questions before the
meeting. In creating the list of questions, they need to start
with the research question.1 If the investigators have an
idea of what statistics to use, then the questions for the statistician
are related to whether the proposed analyses are
appropriate and what other statistics should be considered.
If the investigators have no idea of what statistics to use,
the first question should be what statistics are appropriate
for the research questions being addressed.
Investigators should take advantage of the meeting with
the statistician to find out why the analysis is appropriate
and to increase their knowledge of statistics. They need to
be able to defend their choice of statistics at presentations
and within publications.
Several advantages result from having a plan for data
analysis before starting the study. The most obvious is that
the investigators are not left wondering what to do with all
of the data they now have in their computer. A plan speeds
the process of data analysis. If a computer program will be
used, the commands for the analysis can even be written
before data collection is complete. In this case, as soon as all
of the data are entered, the investigators run the predetermined
programs, and the analysis is ready for interpretation.
The second advantage of planning the statistical analysis
before the study is an increase in scientific integrity. The
investigators who have a plan ahead of time are less likely
to bend the analysis to suit their purpose. A plan also prevents
the process of repeating analyses until something is
found that is statistically significant. A post hoc (after the
fact) approach to statistical analysis is inappropriate and
increases the chance of making a type I statistical error (see
a future issue of this series on hypothesis testing for a discussion
of type I errors).2 If post hoc analyses are used, a
technique such as Bonferroni adjustment is needed to
decrease the chance of a type 1 error.2
Before beginning a study, the investigators should identify
the computer, the data entry method,3 and the data
analysis software they will use for the study. They also
should spend time during the early phases of the project
becoming familiar with the software to be used. Data analysis
will proceed more smoothly if the investigators do not
need to stop and ask for technical assistance.