Chapter 5 Variables and Hypotheses

The Importance of Studying Relationships

We mentioned in Chapter 2 that an important characteristic of many research questions is that they suggest a
relationship of some sort to be investigated. Not all research questions, however, suggest relationships.
Sometimes researchers are interested only in obtaining descriptive information to fi nd out how people think or
feel or to describe how they behave in a particular situation. Other times the intent is to describe a particular
program or activity. Such questions also are worthy of investigation. As a result, researchers may ask questions
like the following:
• How do the parents of the sophomore class feel about the counseling program?
• What changes would the staff like to see instituted in the curriculum?
• Has the number of students enrolling in college preparatory as compared to noncollege preparatory courses
changed over the last four years?
• How does the new reading program differ from the one used in this district in the past?
• What does an inquiry-oriented social studies teacher do?

Notice that no relationship is suggested in these questions. The researcher simply wants to identify
characteristics, behaviors, feelings, or thoughts. It is often necessary to obtain such information as a first step in
designing other research or making educational decisions of some sort.

The problem with purely descriptive research questions is that answers to them do not help us understand why
people feel or think or behave a certain way, why programs possess certain characteristics, why a particular
strategy is to be used at a certain time, and so forth. We may learn what happened, or where or when (and even
how) something happened, but not why it happened. As a result, our understanding of a situation, group, or
phenomenon is limited. For this reason, scientists highly value research questions that suggest relationships to
be investigated, because the answers to them help explain the nature of the world in which we live. We learn to
understand the world by learning to explain how parts of it are related. We begin to detect patterns or
connections between the parts.

We believe that understanding is generally enhanced by the demonstration of relationships or

connections. It is for this reason that we favor the formation of a hypothesis that predicts the existence of a
relationship. There may be times, however, when a researcher wants to hypothesize that a relationship does not
exist. Why so? The only persuasive argument we know of is that of contradicting an existing widespread (but
perhaps erroneous) belief. For example, if it can be shown that a great many people believe, in the absence of
adequate evidence, that young boys are less sympathetic than young girls, a study in which a researcher fi nds
no difference between boys and girls (i.e., no relationship between gender and sympathy) might be of value
(such a study may have been done, although we are not aware of one). Unfortunately, most (but by no means all)
of the methodological mistakes made in research (such as using inadequate instruments or too small a sample of
participants) increase the chance of fi nding no relationship between variables. (We shall discuss several such
mistakes in later chapters.)

Variables
WHAT IS A VARIABLE?

At this point, it is important to introduce the idea of variables, since a relationship is a statement about
variables. What is a variable? A variable is a concept—a noun that stands for variation within a class of objects,
such as chair, gender, eye color, achievement, motivation, or running speed. Even spunk, style, and lust for life
are variables. Notice that the individual members in the class of objects, however, must differ—or vary—to
qualify the class as a variable. If all members of a class are identical, we do not have a variable. Such
characteristics are called constants , since the individual members of the class are not allowed to vary, but rather
are held constant. In any study, some characteristics will be variables, while others will be constants. Suppose a
researcher is interested in studying the effects of reinforcement on student achievementrade level of the students
would be a constant.
Notice that it is easier to see what some of these concepts stand for than others. The concept of chair,
for example, stands for the many different objects that we sit on that possess legs, a seat, and a back.
Furthermore, different observers would probably agree as to how particular chairs differ. It is not so easy,
however, to see what a concept like motivation stands for, or to agree on what it means. The researchers must be
specific here—they must defi ne motivation as clearly as possible. They must do this so that it can be measured
or manipulated. We cannot meaningfully measure or manipulate a variable if we cannot defi ne it. As we
mentioned above, much educational research involves looking for a relationship among variables. But what
variables?
There are many variables “out there” in the world that can be investigated. Obviously, we can’t
investigate them all, so we must choose. Researchers choose certain variables to investigate because they
suspect that these variables are somehow related and believe that discovering the nature of this relationship, if
possible, can help us make more sense out of the world in which we live.

QUANTITATIVE VERSUS CATEGORICAL VARIABLES

Variables can be classifi ed in several ways. One way is to distinguish between quantitative and
categorical variables. Quantitative variables exist in some degree (rather than all or none) along a continuum
from less to more, and we can assign numbers to different individuals or objects to indicate how much of the
variable they possess. Two obvious examples are height (John is 6 feet tall and Sally is 5 feet 4 inches) and
weight (Mr. Adams weighs only 150 pounds and his wife 140 pounds, but their son tips the scales at an even
200 pounds). We can also assign numbers to various individuals to indicate how much “interest” they have in a
subject, with a 5 indicating very much interest, a 4 much interest, a 3 some interest, a 2 little interest, a 1 very
little interest, down to a 0 indicating no interest. If we can assign numbers in this way, we have the variable
interest.
 Quantitative variables can often (but not always) be subdivided into smaller and smaller units. Length,
for example, can be measured in miles, yards, feet, inches, or in whatever subdivision of an inch is needed. By
way of contrast, categorical variables do not vary in degree, amount, or quantity but are qualitatively different.
Examples include eye color, gender, religious preference, occupation, position on a baseball team, and most
kinds of research “treatments” or “methods.” For example, suppose a researcher wishes to compare certain
attitudes in two different groups of voters, one in which each individual is registered as a member of one
political party and the other in which individuals are members of another party. The variable involved would be
political party. This is a categorical variable—a person is either in one or the other category, not somewhere in
between being a registered member of one party and being a registered member of another party. All members
within each category of this variable are considered the same as far as party membership is concerned (see
Figure 5.1).
Can teaching method be considered a variable? Yes, it can. Suppose a researcher is interested in
studying teachers who use different methods in teaching. The researcher locates one teacher who lectures
exclusively, another who buttresses her lectures with slides, films, and computer images, and a third who uses
the case-study method and lectures not at all. Does the teaching method “vary”? It does. You may need to
practice thinking of differences in methods or in groups of people (teachers compared to administrators, for
example) as variables, but mastering this idea is extremely useful in learning about research.
Now, here are several variables. Which ones are quantitative variables and which ones are categorical variables?
1. Make of automobile
2. Learning ability
3. Ethnicity
4. Cohesiveness
5. Heartbeat rate
6. Gender *
Researchers in education often study the relationship between (or among) either (1) two (or more) quantitative
variables; (2) one categorical and one quantitative variable; or (3) two or more categorical variables. Here are
some examples of each:
1. Two quantitative variables
 Age and amount of interest in school
 Reading achievement and mathematics achievement
 Classroom humanism and student motivation
 Amount of time watching television and aggressiveness of behavior
2. One categorical and one quantitative variable
Method used to teach reading and reading achievement
• Counseling approach and level of anxiety
• Nationality and liking for school
• Student gender and amount of praise given by teachers
3. Two categorical variables
• Ethnicity and father’s occupation
• Gender of teacher and subject taught
 • Administrative style and college major
• Religious affi liation and political party membership

Sometimes researchers have a choice of whether to treat a variable as quantitative or categorical. It is not
uncommon, for example, to fi nd studies in which a variable such as anxiety is studied by comparing a group of
“high-anxiety” students to a group of “low-anxiety” students. This treats anxiety as though it were a categorical
variable. While there is nothing really wrong with doing this, there are three reasons why it is preferable in such
situations to treat the variable as quantitative.
1. Conceptually, we consider variables such as anxiety in people to be a matter of degree, not a matter of
either-or.
2. Collapsing the variable into two (or even several) categories eliminates the possibility of using more
detailed information about the variable, since differences among individuals within a category are
ignored.
3. The dividing line between groups (for example, between individuals of high, middle, and low anxiety)
is almost always arbitrary (that is, lacking in any defensible rationale).

INDEPENDENT VERSUS
DEPENDENT VARIABLES
A common and useful way to think about variables is to classify them as independent or dependent.
Independent variables are those that the researcher chooses to study in order to assess their possible effect(s)
on one or more other variables. An independent variable is presumed to affect (at least partly cause) or somehow
influence at least one other variable. The variable that the independent variable is presumed to affect is called a
dependent variable . In commonsense terms, the dependent variable “depends on” what the independent
variable does to it, how it affects it. For example, a researcher studying the relationship between childhood
success in mathematics and adult career choice is likely to refer to the former as the independent variable and
subsequent career choice as the dependent variable.
It is possible to investigate more than one independent (and also more than one dependent) variable in a study.
For simplicity’s sake, however, we present examples in which only one independent and one
dependent variable are involved. The relationship between independent and dependent
variables can be portrayed graphically as follows:
Affects
Independent Dependent
variable(s) variable(s)

(presumed or (presumed
possible cause) results)

At this point, let’s check your understanding. Suppose a researcher plans to investigate the following
question: “Will students who are taught by a team of three teachers learn more science than students taught by
one individual teacher?” What are the independent and dependent variables in this question? *
 Notice that there are two conditions (sometimes called levels ) in the independent variable—“three
teachers” and “one teacher.” Also notice that the dependent variable is not “science learning” but “ amount of
science learning.” Can you see why?
At this point, things begin to get a bit complicated. Independent variables may be either manipulated or
selected. A m anipulated variable is one that the researcher creates. Such variables are typically found in
experimental studies (see Chapter 13). Suppose, for example, that a researcher decides to investigate the effect
of different amounts of reinforcement on reading achievement and systematically assigns students to three
different groups. One group is praised continuously every day during their reading session; the second group is
told simply to “keep up the good work”; the third group is told nothing at all. The researcher, in effect,
manipulates the conditions in this experiment, thereby creating the variable amount of reinforcement. Whenever
a researcher sets up experimental conditions, one or more variables are created. Such variables are called
manipulated variables, experimental variables , or treatment variables .
Sometimes researchers select an independent variable that already exists. In this case, the researcher
must locate and select examples of it, rather than creating it. In our earlier example of reading methods, the
researcher would have to locate and select existing examples of each reading method, rather than arranging for
them to happen. Selected independent variables are not limited to studies that compare different treatments; they
are found in both causal-comparative and correlational studies (see Chapters 15 and 16). They can be either
categorical or quantitative. The key idea here, however, is that the independent variable (either created or
selected) is thought to affect the dependent variable. Here are a few examples of some possible relationships
between a selected independent variable and a dependent variable:

Independent Variable Dependent Variable

Gender (categorical) Musical aptitude (quantitative)
Mathematical ability (quantitative) Career choice (categorical)
Gang membership (categorical) Subsequent marital status (categorical)
Test anxiety (quantitative) Test performance (quantitative)

Notice that none of the independent variables in the above pairs could be directly manipulated by the
researcher. Notice also that, in some instances, the independent/ dependent relationship might be reversed,
depending on which one the researcher thought might be the cause of the other. For example, he or she might
think that test performance causes anxiety, not the reverse.
Generally speaking, most studies in education that have one quantitative and one categorical variable
are studies comparing different methods or treatments. As we indicated above, the independent variable in such
studies (the different methods or treatments) represents a categorical variable. Often the other (dependent)
variable is quantitative and is referred to as an outcome variable . * The reason is rather clear-cut. The
investigator, after all, is interested in the effect(s) of the differences in method on one or more outcomes (student
achievement, their motivation, interest, and so on). Again, let’s check your understanding. Suppose a researcher
plans to investigate the following question: “Will students like history more if taught by the inquiry method than
if taught by the case-study method?” What is the outcome variable in this question?

MODERATOR VARIABLES
 A moderator variable is a special type of independent variable. It is a secondary independent variable
that has been selected for study in order to determine if it affects or modifi es the basic relationship between the
primary independent variable and the dependent variable. Thus, if an experimenter thinks that the relationship
between variables X and Y might be altered in some way by a third variable Z, then Z could be included in the
study as a moderator variable.
Consider an example. Suppose a researcher is interested in comparing the effectiveness of a discussion-
oriented approach to a more visually oriented approach for teaching a unit in a U.S. History class. Suppose
further that the researcher suspects that the discussion approach may be superior for the girls in the class (who
appear to be more verbal and to learn better through conversing with others) and that the visual approach may be
more effective for boys (who seem to perk up every time a video is shown). When the students are tested
together at the end of the unit, the overall results of the two approaches may show no difference, but when the
results of the girls are separated from those of the boys, the two approaches may reveal different results for each
subgroup. If so, then the gender variable moderates the relationship between the instructional approach (the
independent variable) and effectiveness (the dependent variable). The infl uence of this moderator variable can
be seen in Figure 5.2.
Here are two examples of research questions that contain moderator variables.
Research Question 1: “Does anxiety affect test performance and, if so, does it depend on test-taking
experience?”
• Independent variable: anxiety level
• Moderator variable: test-taking experience
• Dependent variable: test performance
Research Question 2: “Do high school students taught primarily by the inquiry method perform better on tests
of critical thinking than high school students taught primarily by the demonstration method and, if so, does it
vary with grade level?”
• Independent variable: instructional method
• Moderator variable: grade level
• Dependent variable: performance on critical thinking tests
As you can see, the inclusion of a moderator variable (or even two or three) in a study can provide
considerably more information than just studying a single independent variable alone. We recommend their
inclusion whenever appropriate.

MEDIATOR VARIABLES
While a moderator variable can modify or infl uence the strength of a relationship between two other variables,
a mediator variable is one that attempts to explain the relationship between the two other variables. Let us
reexamine the relationship in Hypothesis 1 above between anxiety level (AL) and test performance (TP) on a
high-stakes test like the SAT exam. The moderator variable in this case is test-taking experience (TTE) because
the relationship between AL and TP is stronger for students with prior experience taking the SAT. A possible
mediator variable in this case could be socioeconomic status (SES) because it could explain why there is a
relationship between AL and TP.
EXTRANEOUS VARIABLES
A basic problem in research is that there are many possible independent variables that could have an
effect on the dependent variables. Once researchers have decided which variables to study, they must be
concerned about the infl uence or effect of other variables that exist. Such variables are usually called
extraneous variables. The task is to control these extraneous variables somehow to eliminate or minimize their
effect.
Extraneous variables are independent variables that have not been controlled. Look again at the
research question about team teaching on page 80. What are some other variables that could have an effect on
the learning of students in a classroom situation? There are many possible extraneous variables. The personality
of the teachers involved, the experience level of the students, the time of day the classes are taught, the nature of
the subject taught, the textbooks used, the type of learning activities the teachers employ, and the teaching
methods—all are possible extraneous variables that could affect learning in this study. Figure 5.3 illustrates the
importance of identifying extraneous variables.

The principal of a high school compares the final examination scores of two history classes taught by teachers who use different methods,
not realizing that they are also different in many other ways because of extraneous variables. The classes differ in:
Extraneous variables : • Size of class
• Gender of students
• Gender of teacher
• Age of teacher
• Time of day class meets
• Days of week class meets
• Ethnicity of teacher
• Length of class

One way to control extraneous variables is to hold them constant. For example, if a researcher includes
only boys as the subjects of a study, she is controlling the variable of gender. We would say that the gender of
the subjects does not vary; it is a constant in this study. Researchers must continually think about how they
might control the possible effect(s) of extraneous variables. We will discuss how to do this in some detail in
Chapter 9, but for now you need to make sure you understand the difference between independent and
dependent variables and to be aware of extraneous variables. Try your hand at the following question: “Will
female students who are taught history by a teacher of the same gender like the subject more than female
students taught by a teacher of a different gender?” What are the variables?

Hypotheses
WHAT IS A HYPOTHESIS?
A hypothesis is, simply put, a prediction of the possible outcomes of a study. For example, here is a
research question followed by its restatement in the form of a possible hypothesis:
Question: Will students who are taught history by a teacher of the same gender like the subject more than
students taught by a teacher of a different gender?
Hypothesis: Students taught history by a teacher of the same gender will like the subject more than students
taught history by a teacher of a different gender.
Here are two more examples of research questions followed by the restatement of each as a possible
hypothesis:
Question: Is rapport with clients of counselors using client-centered therapy different from that of counselors
using behavior-modifi cation therapy?

*The dependent variable is liking for history, the independent variable is the gender of the teacher. Possible extraneous variables
include the personality and ability of the teacher(s) involved; the personality and ability level of the students; the materials used, such as
textbooks; the style of teaching; ethnicity and/or age of the teacher and students; and others. The researcher would want to control as many
of these variables as possible.

Hypothesis: Counselors who use a client-centered therapy approach will have a greater rapport with their
clients than counselors who use a behaviormodifi cation approach.
Question: How do teachers feel about special classes for the educationally handicapped?
Hypothesis: Teachers in XYZ School District believe that students attending special classes for the
educationally handicapped will be stigmatized.
or
Teachers in XYZ School District believe that special classes for the educationally handicapped will help such
students improve their academic skills.

ADVANTAGES OF STATING HYPOTHESES IN ADDITION TO RESEARCH QUESTIONS

Stating hypotheses has both advantages and disadvantages. What are some of the advantages? First, a
hypothesis forces us to think more deeply and specifi cally about the possible outcomes of a study. Elaborating
on a question by formulating a hypothesis can lead to a more sophisticated understanding of what the question
implies and exactly what variables are involved. Often, as in the case of the third example above, when more
than one hypothesis seems to suggest itself, we are forced to think more carefully about what we really want to
investigate.
A second advantage of restating questions as hypotheses involves a philosophy of science. The
rationale underlying this philosophy is as follows: If one is attempting to build a body of knowledge in addition
to answering a specifi c question, then stating hypotheses is a good strategy because it enables one to make
specific predictions based on prior evidence or theoretical argument. If these predictions are borne out by
subsequent research, the entire procedure gains both in persuasiveness and effi ciency. A classic example is
Albert Einstein’s theory of relativity. Many hypotheses were formulated as a result of Einstein’s theory, which
were later verifi ed through research. As more and more of these predictions were shown to be fact, not only did
they become useful in their own right, they also provided increasing support for the original ideas in Einstein’s
theory, which generated the hypotheses in the fi rst place. Lastly, stating a hypothesis helps us see if we are, or
are not, investigating a relationship. If not, we may be prompted to formulate one.

DISADVANTAGES OF STATING HYPOTHESES

Essentially, the disadvantages of stating hypotheses are threefold. First, stating a hypothesis may lead
to a bias , either conscious or unconscious, on the part of the researcher. Once investigators state a hypothesis,
they may be tempted to arrange the procedures or manipulate the data in such a way as to bring about a desired
outcome.
This is probably more the exception than the rule. Researchers are assumed to be intellectually honest
—although there are some famous exceptions. All studies should be subject to peer review; in the past, a review
of suspect research has, on occasion, revealed such inadequacies of method that the reported results were cast
into doubt. Furthermore, any particular study can be replicated to verify the fi ndings of the study.
Unfortunately, few educational research studies are repeated, so this “protection” is somewhat of an illusion. A
dishonest investigator stands a fair chance of getting away with falsifying results. Why would a person
deliberately distort his or her fi ndings? Probably because professional recognition and fi nancial reward accrue
to those who publish important results. Even for the great majority of researchers who are honest, however,
commitment to a hypothesis may lead to distortions that are unintentional and unconscious. But it is probably
unlikely that any researcher in the field of education is ever totally disinterested in the outcomes of a study;
therefore, his or her attitudes and/or knowledge may favor a particular result. For this reason, we think it is
desirable for researchers to make known their predilections regarding a hypothesis so that they are clear to
others interested in their research. This also allows investigators to take steps to guard (as much as possible)
against their personal biases.
The second disadvantage of stating hypotheses at the outset is that it may sometimes be unnecessary, or
even inappropriate, in research projects of certain types, such as descriptive surveys and ethnographic studies. In
many such studies, it would be unduly presumptuous, as well as futile, to predict what the fi ndings of the
inquiry will be.
The third disadvantage of stating hypotheses is that focusing attention on a hypothesis may prevent
researchers from noticing other phenomena that might be important to study. For example, deciding to study the
effect of a “humanistic” classroom on student motivation might lead a researcher to overlook its effect on such
characteristics as sex-typing or decision making, which would be quite noticeable to another researcher who was
not focusing solely on motivation. This seems to be a good argument against all research being directed toward
hypothesis testing.
Consider the example of a research question presented earlier in this chapter: “How do teachers feel
about special classes for the educationally handicapped?” We offered two (of many possible) hypotheses that
might arise out of this question: (1) “Teachers believe that students attending special classes for the
educationally handicapped will be stigmatized” and (2) “Teachers believe that special classes for the
educationally handicapped will help such students improve their academic skills.” Both of these hypotheses
implicitly suggest a comparison between special classes for the educationally handicapped and some other kind
of arrangement. Thus, the relationship to be investigated is between teacher beliefs and type of class. Notice that
it is important to compare what teachers think about special classes with their beliefs about other kinds of
arrangements. If researchers looked only at teacher opinions about special classes without also identifying their
views about other kinds of arrangements, they would not know if their beliefs about special classes were in any
way unique or different.

IMPORTANT HYPOTHESES
As we think about possible hypotheses suggested by a research question, we begin to see that some of them are
more important than others. What do we mean by important? Simply that some may lead to more useful
knowledge. Compare, for example, the following pairs of hypotheses. Which hypothesis in each pair would you
say is more important?
Pair 1
a. Second-graders like school less than they like watching television.
b. Second-graders like school less than fi rst-graders but more than third-graders.
Pair 2
a. Most students with academic disabilities prefer being in regular classes rather than in special classes.
b. Students with academic disabilities will have more negative attitudes about themselves if they are placed in
special classes than if they are placed
in regular classes.
Pair 3
a. Counselors who use client-centered therapy procedures get different reactions from counselees
than do counselors who use traditional therapy procedures.
b. Counselees who receive client-centered therapy express more satisfaction with the counseling process
than do counselees who receive traditional therapy.

In each of the three pairs, we think that the second hypothesis is more important than the fi rst, since in each
case (in our judgment) not only is the relationship to be investigated clearer and more specifi c but also
investigation of the hypothesis seems more likely to lead to a greater amount of knowledge. It also seems to us
that the information to be obtained will be of more use to people interested in the research question.

DIRECTIONAL VERSUS
NONDIRECTIONAL HYPOTHESES
Let us make a distinction between directional and nondirectional hypotheses. A directional hypothesis
indicates the specifi c direction (such as higher, lower, more, or less) that a researcher expects to emerge in a
relationship.
The particular direction expected is based on what the researcher has found in the literature, in theory,
or from personal experience. The second hypothesis in each of the three pairs above is a directional hypothesis.
Sometimes it is diffi cult to make specifi c predictions. If a researcher suspects that a relationship exists but has
no basis for predicting the direction of the relationship, she cannot make a directional hypothesis. A
nondirectional hypothesis does not make a specific prediction about what direction the outcome of a study will
take. In nondirectional form, the second hypotheses of the three pairs above would be stated as follows:
1. First-, second-, and third-graders will feel differently toward school.
2. There will be a difference between the scores on an attitude measure of students with academic disabilities
placed in special classes and such students placed in regular classes.
3. There will be a difference in expression of satisfaction with the counseling process between counselees
who receive client-centered therapy and counselees who receive traditional therapy.
Figure 5.5 illustrates the difference between a directional and a nondirectional hypothesis. If the person pictured
is approaching a street corner, three possibilities exist when he reaches the corner:
• He will continue to look straight ahead.
• He will look to his right.
• He will look to his left.
A nondirectional hypothesis would predict that he will look one way or the other. A directional
hypothesis would predict that he will look in a particular direction (for example, to his right). Since a directional
hypothesis is riskier (because it is less likely to occur), it is more convincing when confirmed.
Both directional and nondirectional hypotheses appear in the literature of research, and you should
learn to recognize each.

HYPOTHESES AND QUALITATIVE RESEARCH

What is notable about the formation of hypotheses in qualitative research is that they are typically not
stated at the beginning of a study, but rather they emerge as a study progresses. Rather than testing hypotheses
as in quantitative studies, qualitative researchers are more likely to generate new hypotheses as a result of what
they find as they go about their work—as they observe patterns and relationships in the natural setting rather
than hypothesizing what such patterns and relationships might be beforehand. Many qualitative researchers do
state some of their ideas before they begin a study, but these are usually called propositions rather than
hypotheses. Propositions differ from hypotheses in that they are not intended to be tested against the data (as in
quantitative research) but rather are viewed as flexible tools intended to help guide researchers in their collection
and analysis of qualitative data. The reluctance of qualitative researchers to formulate hypotheses at the
beginning of a study is based on their conviction that participants and situations often differ widely and must
first be understood before any hypotheses can be suggested.

1. Here are several research questions. Which ones suggest relationships?

a) How many students are enrolled in the sophomore class this year?
b) As the reading level of a text passage increases, does the number of errors
students make in pronouncing words in the passage increase?
c) Do individuals who see themselves as socially “attractive” expect their romantic
partners also to be (as judged by others) socially attractive?
d) What does the faculty dislike about the new English curriculum?
e) Who is the brightest student in the senior class?
f) Will students who score above the 90th percentile on a standardized reading test
also score above the 90th percentile on a standardized writing test?
g) Which political party contains the most Protestants—Democratic or Republican?
2. How would you rank the questions in item 1 in terms of signifi cance? Why?
3. What might cause a researcher to state a directional hypothesis rather than a
nondirectional hypothesis? What about the reverse?
4. Are there any variables that researchers should not study? Explain.
5. It is often argued that we cannot meaningfully measure a variable if we cannot defi ne
it. Is this true? always? Discuss.
6. “Commitment to a hypothesis may lead to distortions that are unintentional and
unconscious.” Would you agree? If so, can you give an example of such a hypothesis?
7. Can you think of a possible study for which it would be presumptuous to predict the
outcome?