ANALYSIS OF QUANTITATIVE DATA
EXPANSION BOX 4
Statistical Programs on Computers
Almost every social researcher who needs to calcu-
late many statistics does so with a computer program.
One can calculate some statistics using a basic spread-
sheet program, such as Excel. Unfortunately, spread-
sheets are designed for accounting and bookkeeping
functions; they include statistical functions but are
clumsy and limited for that purpose. There are many
computer programs designed for calculating general
statistics. The marketplace can be confusing to a
beginner for products rapidly evolve with changing
computer technology. One or two decades ago, one
had to know a computer language or do simple pro-
gramming to have a computer calculate statistics.
In recent years, the software has become less
demanding for a user. The most popular programs in
the social sciences are Minitab, Microcase, and Stas-
tical Package for the Social Sciences (SPSS). Others
include Statistical Analysis System (SAS), BMPD
(bought by SPSS, Inc.), STATISTICA by StratSoft, and
Strata. Many began as simple, low-cost programs for
research purposes. Today private corporations own
many of these and are interested in selling a sophis-
ticated set of software products to many diverse cor-
porate and government users.
The most widely used program for statistics in the
social sciences is SPSS. Its advantages are that social
researchers have used it extensively for more than
three decades, it includes many ways to manipulate
quantitative data, and it contains most statistical mea-
sures. Its disadvantage is that it can take a long time
to learn because of its many options and complex
statistics. Also, it is expensive to purchase except for
an inexpensive, “stripped down” student version
included with a textbook or workbook.
As computer technology makes using statistics
programs easier, the danger increases that some
people will use the programs but not understand sta-
tistics or what the programs are doing. These people
can easily violate basic assumptions required by a
statistical procedure, use the statistics improperly,
and produce results that are pure nonsense yet look
very technically sophisticated.
support a hypothesis. There is nothing wrong with
rejecting a hypothesis. The goal of scientific
research is to produce knowledge that truly reflects
the social world, not to defend pet ideas or hypothe-
ses. Hypotheses are theoretical guesses based
on limited knowledge; they need to be tested.
Excellent-quality research can find that a hypoth-
esis is wrong, and poor-quality research can sup-
port a hypothesis. Good research depends on
high-quality methodology, not on supporting a spe-
cific hypothesis.
Good research means guarding against pos-
sible errors or obstacles to true inferences from data
to the social world. Errors can enter into the research
process and affect results at many places: research
design, measurement, data collection, coding, cal-
culating statistics and constructing tables, or inter-
preting results. Even if you can design, measure,
collect, code, and calculate without error, you must
also complete another step in the research process:
interpret the tables, charts, and statistics, and answer
the question: What does it all mean? The only way
to assign meaning to facts, charts, tables, or statis-
tics is to use theory, insight, and understanding.
Data, tables, or computer output alone cannot
answer research questions. The facts do not speak
for themselves. As a researcher, you must return to
your theory (i.e., concepts, relationships among
concepts, assumptions, theoretical definitions) and
give the results meaning. Do not lock yourself into
the ideas with which you began. There is room for
creativity, and new ideas are generated by trying to
figure out what results really say. It is important to
be careful in designing and conducting research so
that you can look at the results as a reflection of
something in the social world and not worry about
whether they are due to an error or an artifact of the
research process itself.
Before we leave quantitative research, we must
present one last issue. Journalists, politicians, and
others increasingly use statistical results to make a
point or bolster an argument. This has not produced
