Ninth Edition CoverGraziano & Raulin
Research Methods (9th edition)

Chapter 8 Fill-in-the-Blank Questions
Hypothesis Testing, Validity,
and Threats to Validity

Now that you have familiarity with the new key terms of this chapter, it is time to begin using the terms in the context of the chapter material. See if you can fill-in-the-blanks in the following discussions. To get the correct answer, click on the blank.

  1. The testing of hypotheses is a critical part of experimental research. Each experiment tests at least one __________-__________ . The research hypothesis may begin as vague or general ideas, which are refined into a(n) __________-__________ -__________-__________. The statement of the problem suggests the expected relationship between __________. It is converted into a specific research hypothesis by __________- __________ the research variables.

  2. Statements of the problem are usually in the form of __________ that concern the causal relationship between two variables. The statement of the problem should always define (1) the expected causal __________, (2) the __________ to be studied, and (3) an indication of the expected __________ of the causal effect.

  3. A(n) __________-__________ of a variable defines precisely how the variable will be measured or manipulated. Variables are usually __________, which are not directly observable, but inferred by the researcher. In most cases, these constructs can be operationalized in many different ways. For example, one could measure anxiety by observing and counting anxious behavior, measuring physiological changes, or asking people how anxious they feel. Each of the above would represent a(n) __________- __________ of the construct of anxiety.

  4. The __________-__________ restates the statement of the problem by defining each of the variables in operational terms. Since __________ can often be defined operationally in many different ways, several __________-__________ can be generated from a single statement of the problem. Further, the wording of the research hypothesis can reflect a specific research design to be used. A critical role in developing research hypotheses is played by __________, which guides the thinking of the researcher.

  5. Testing the __________-__________ actually involves testing three hypotheses. The first is the __________ or statistical hypothesis. This hypothesis evaluates whether the observed relationship between variables is strong enough that it is unlikely to be a chance finding. The second hypothesis is the __________-__________ hypothesis. The confounding variable hypothesis is actually an entire set of hypotheses. Each confounding variable hypothesis states that the observed relationship between __________ and __________ variables may actually be the result of some other __________ variable. For example, if a store institutes a very visible anti-shoplifting publicity campaign and observes a significant reduction in shoplifting, a possible confounding variable hypothesis is that the reduction is not due to the campaign, but rather to increased vigilance on the part of the sales personnel. Unless confounding variable hypotheses can be __________-__________, we cannot confidently draw a causal conclusion from a research study. The third hypothesis is the __________ hypothesis, which states that the __________ variable has the predicted causal effect on the __________ variable. We can accept this last hypothesis only if the observed relationship is strong enough that we can reject the __________ hypothesis and the experimental design provides sufficient control that we can rule out all __________-__________ hypotheses.

  6. A major concern in research is the validity of the procedures we use and the conclusions we draw. We make the distinction between several types of validity. __________ validity is concerned with the accuracy of the statistical conclusions. It might be threatened, for example, if the measure of the dependent variable is __________ or if we violate a(n) __________ of the statistical procedure used to test the null hypothesis. __________ validity refers to how well the study's results support the theory or constructs behind the research. If another theory provides a better theoretical explanation of the results, the construct validity would be in doubt. __________ validity refers to the degree to which we are able to generalize the results of the study to other participants, conditions, times, and places. External validity is threatened if the __________ does not adequately represent the participants, conditions, times, or places to which we wish to generalize. __________ validity refers to how confidently we can infer a causal connection between the __________ and __________ variables. Internal validity is threatened whenever __________-__________ hypotheses cannot be ruled out.

  7. Although, in theory, there are an unlimited number of possible confounding variables, a few major confounding variables represent most of the threats to __________ validity. __________ refers to the natural changes that occur over time. It is particularly important when studying children, because children can show rapid maturational changes. __________ refers to any factor that occurs during the time between pretest and posttest measurements that might affect the dependent measure. __________ refers to changes that the participant shows between a pretest and posttest that is the result of increasing proficiency in the testing situation. __________ refers to the changes that might occur in how the dependent measure is obtained at two points in time. Instrumentation is a particularly important problem if the measuring instrument is a human observer, who might change criteria over time without being aware of it. __________- __________-__________-__________ refers to the idea that, when participants are specifically selected because they are at the extreme of the distribution, on any retest we can expect the participants to be less extreme on average. __________ occurs when the groups to be studied are created in such a manner that they are not comparable at the beginning of the study. Hence, any observed differences between the groups at the end of the study might actually reflect initial differences between the groups instead of the effects of the independent variable on the dependent variable. __________ refers to participants dropping out of the study before its completion. When participants drop out of a study, the possibility exists that the participants who remain are no longer representative of the original group. __________-__________-__________ occurs when the differences between experimental conditions are smaller than the researcher intended because participants become aware of the procedures in the other conditions. This is especially likely when potential participants are in close contact with one another. Finally, __________-__________ occurs when participants are tested under more than one experimental condition. If the effects of one condition carry over and affect performance in the conditions that follow, then potential confounding exists.

  8. In addition to the confounding variables discussed above, validity can be threatened by other factors. __________-__________ refers to any change in participants' behavior as a result of the participants knowing that they are being studied. Participants are likely to behave differently when they know that they are being observed. Participants often look for cues in a study that will indicate how a participant is expected to behave. These cues are sometimes referred to as __________-__________ and may be unintentionally transmitted to the participant by the researcher. A related phenomenon is the __________-__________, which occurs when participants expect some fairly specific effect of an experimental manipulation. For example, if a participant expects some improvement as the result of a treatment, such improvement is often observed, even when the treatment itself is actually useless. The observed improvement seems to be due to the suggestion that the treatment will be effective, rather than to the actual value of the treatment.

  9. There is another class of confounding variables known as __________ effects, which are a large class of behaviors that can directly or indirectly affect the findings of a study. Experimenter effects are usually based on __________- __________, which can cause a researcher to bias the results of a study in several ways. For example, the researcher might influence the participant to __________ in a manner that supports the hypotheses or might select data or a statistical technique that highlights the expected effect. The researcher is usually not __________ of these biases. Consequently, these effects can be very difficult to control.

  10. Even if _________ significant differences in the _________ direction are found, we still _________ be sure that the observed differences are due to the _________ _________. They might be due to _________ _________. Rejecting the _________ _________ is necessary, but not _________, to draw a _________ _________. We must also rule out confounding _________ as explanations of the _________. It is best to do this during the _________ phase, when you anticipate the confounding variables and include _________ to eliminate their effects.