# 2: HRAF
Data Recovery (requires use of the statistical program SPSS)
by Brad Huber, Department of Sociology and
Anthropology, College of Charleston
DUE:__________________________, at the
beginning of class
A project you will be working on in the future, Project #3 (see Anthropological
Research Methods--Project #3: Group Poster Presentation),
will have you analyze cross-cultural data using SPSS 10 and
report your findings in the form of a research poster. Before
you can undertake this project, you need to develop a
cross-cultural data set. That is the main goal of this project.
In order to attain this goal, you first need to generate
hypotheses that can be tested with cross-cultural data, and
develop conceptual and operational definitions of the
Part A: Familiarize yourselves with a pre-existing data
You will work in pairs. Start by looking at the following
cross-cultural data set called the Standard
Sample-Probability Sample Data, Revised (see * below). A
link to this data set is found by accessing the syllabus for
this course from my website (http://www.cofc.edu/~huberb/)
at College of Charleston. This is an SPSS 10 data file. You'll
be able to open this link because our campus' computer labs have
SPSS 10 installed on them.
Note by the editors: The following instructions are
specifically for College of Charleston students. Performing
these exercises requires the statistical software program SPSS
version 10 for other users.
You will see that this data set has lots of variables already
coded. To see that these variables deal with different
characteristics of cultures, select the "Variable
View" tab at the bottom of the page, and look at the
"labels" and "values". An even better way to
look at all of the variables is to return to the "Data
View" and select "Utilities" and "File
Information". (After you do that, I suggest you right click
on the "List of Variables on the working file", and
scroll down to the "SPSSrtf document object", and then
click "open". You should be able to scroll all the way
down to Variable O914 if you want to. I caution you against
printing out the entire list of variables since that would use
up a lot of paper. For your convenience, you can look at the
variable list by clicking a link found on my online syllabus
called, "Standard Sample-Probability Sample, Revised
Now that you have familiarized yourselves with this data set,
randomly select 30 of 39 cultures. Then delete the 9 cultures
that were not randomly selected. Afterwards, save the file under
a new name like, "PROJECT3.sav". Eventually you will
be adding 4 of your own variables to this data set, but that
comes later, in Project #3.
*What is a data set? In a nutshell,
it contains ethnographic data of all sorts for the 39 societies
that are found in both the Standard Cross Cultural Sample and
the Probability Sample. No variable in this data set has
more than 30% of its cases missing.
Part B: Developing four (4) cross-cultural hypotheses:
Developing four(4)cross-cultural hypotheses requires you to
think critically. When you are developing them, I would like you
to do the following:
1) All of your hypotheses should deal with different aspects of
the same topic, e.g., different aspects of marriage or food
production or medicinal plant use. Select a topic that you are
very interested it.
2) Each of the four hypotheses should contain two (2) and only
two variables. One of the variables in each hypothesis should
come from those that are already in the "Standard Sample -
Probability Sample data" file. You create the other
variable. It should deal with an aspect of a culture that
interests you. See also Basic
Guide to Cross-Cultural Research.
Part C: The project:
Your project consists of you and your partner jointly turning in
Parts 1-6 below. One copy of your work with both of your names
on it is sufficient. Parts 1, 2, 3, and 5 must be typed and
double-spaced. You can use a pen or pencil for Parts 4 and 6.
1) Clearly state four (4) RELATED hypotheses.
2) Give conceptual ("dictionary") definitions
of each of your eight (8) variables. Four variables will come
from the pre-existing data set. The other four (4) will be
created by you.
3) Give operational definitions of the four (4) variables
you created. Some things to keep in mind are:
a) You will use the
information you collect from the eHRAF to assign values to your
variables for each of the 30 societies in your sample. See pp.
38-40 of Bernard's book titled Social Research Methods
text for ideas.
b) Some concepts are
much more difficult to operationalize than others using
ethnographic data. This is because ethnographers do not attend
to some subjects at all or because they pay little attention to
some topics, e.g., few ethnographies contain information on the
pH of the soil.
c) Concepts may be
difficult to operationalize because they are quite abstract,
like the concepts of "community solidarity" or the
"relative status of women". My advice is for you to
develop definitions for specific variables. For example,
instead of "relative status of women" look at
"number of political offices women are reported to
d) Designing operational
definitions requires some trial-and-error. You can expect to
develop a tentative definition and discover that you can't find
that kind of information in the eHRAF files. This means that you
will need to revise your definition.
For a discussion of operational definitions see Bernard's book
titled Social Research Methods and the Basic
Guide to Cross-Cultural Research.
4) Systematically collect and record information for the
four (4) variables you created using a code sheet similar to the
one in the sample project. Then use this information to assign
values/codes to all 30 cultures for all four (4) variables. This
is a very time consuming process but it is made easier if the
four variables you selected are very specific.
5) Now that you've come up with operational definitions
for your variables, and have collected and recorded information
on them, be sure to indicate whether they are at the nominal,
ordinal, interval or ratio level of measurement.
6) Using a chart, enter the values for all four of
the variables you created for all 30 of the societies you
randomly selected. See the example of a chart at the end of the
project. Values can be either letters or numbers, but I very
strongly suggest that you use NUMBERS. Values are ideally
assigned on the basis of good, clear, unambiguous operational
Bernard, H. Russell
1999 Social Research Methods: Qualitative and Quantitative
Approaches. Sage Publications Incorporated.
Check to see how much variability is in your variables. Your
values should not all be the same. If there is little variability,
this is reason for concern. Come talk to me if you encounter
You will be evaluated on the following:
____ 1) 4 Hypotheses
____ 2) Conceptual Definitions
____ 3) Operational Definitions
____ 4) Data on Variables
____ 5) Level of Measurement
____ 6) Chart with Values
Other useful information:
Accessing the eHRAF (and other online databases) from an
Off-Campus Computer at College of Charleston (CofC).
1) Go to the CofC Homepage
2) Then click Library
3) Then Off-Campus Access
5) You will be asked to type in your:
LOGIN ID (Your Social Security
6) After you have logged on, scroll down and select Human
Relations Area Files
View the eHRAF User Guide
on how to use the database.
Making the Font And/Or Icons larger on your computer screen:
1) If you are using Microsoft Word, Go to Format,
then Font. Change the font size to 12pt-16pt.
Click on Default Font, and Say "Yes".
2) If you are using Internet Explorer, Go to View,
then Text Size, and Click on Larger or Largest.
3) If you are using Netscape Communicator, Go to View
and Select Increase Font, or simply hold down
simultaneously Ctrl + [
4)When Using SPSS, Change the Icon Size by clicking View,
ToolBar, Large Buttons. Change font size of the
text in the Data View by clicking View, Font, and Size.