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

Chapter 13 Fill-in-the-Blank Questions
A Second Look at Field Research:
Field Experiments, Program Evaluation,
and Survey Research

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. Research in natural settings is sometimes referred to as __________-__________ . Unlike the naturalistic research discussed in Chapter 6, Chapter 13 discusses experiments in natural settings, which are designed to draw __________ inferences.

  2. Conducting research in natural settings can be difficult. The characteristics of the natural setting often make it difficult for the researcher to __________  all of the sources of extraneous variance. In addition, __________ of the independent variable is often difficult or impossible in the field.

  3. There are three major reasons for conducting experiments in the field. The first is to test the __________ validity of conclusions based on laboratory research. The second is to study the effects of events that occur only in the __________. Finally, testing hypotheses in many different field settings helps to improve the __________ across settings.

  4. The highest degree of control is obtained through __________ , which allow us to draw causal inferences with confidence. However, there are situations in which experiments are impossible. In such situations, a(n) __________-__________ design, which is similar to an experiment in most respects, can provide a reasonable test of a(n) __________ hypothesis. Quasi-experiments control for some __________, but do not have as much control as experiments.

  5. Quasi-experimental designs are similar in form to __________, since a comparison of at least two levels of the __________ variable is made, although the actual __________ is not always under the control of the experimenter. Two quasi-experimental designs were discussed in Chapter 13: the __________ control-group design, and the interrupted __________-__________ design.

  6. A crucial feature of experimental design is the __________ of the groups at the beginning of this study. Normally, this equivalence is established by __________ assigning participants to the groups or conditions of the experiment. In a(n) __________ control-group design, preexisting groups are compared on the dependent measure. The possibility exists that the groups may differ at the start of the study on the __________ variable or may differ on some other variables that could affect the dependent variable. The more __________ these naturally occurring groups are to one another, the closer this design approximates a(n) __________. However, even when the groups are __________ on the dependent variable, it may be possible to make meaningful comparisons.

  7. There are two problems with nonequivalent control-group designs. The first problem is that the groups may be __________ on the dependent measure at the start of the study. The second problem is that there may be other __________ between the groups. To address the possibility that the groups may be different on the dependent measure, a pretest is included. This pretest allows us to evaluate the degree of __________ between the groups prior to the beginning of the study. It also allows us to compute a(n) __________-__________, which shows how much each group changed. In order to control for the possibility that the groups may be different on variables other than the dependent measure, we must first __________ each potential confounding variable and measure each of these variables in order to rule them out.

  8. In an interrupted time-series design, a single group of participants is __________ several times, both before and after some event or __________. The time-series design is an extension of the simple __________-__________ design, but instead of a single pretreatment measure and a single posttreatment measure of the dependent variable, the time-series design uses __________ measures both before and after the manipulation. This series of measures allows us to see trends and may help us to rule out potential __________ variables, such as history, maturation, and regression to the mean.

  9. Some potential __________ variables are well controlled by the interrupted time-series design. By seeing the normal fluctuation over time prior to any manipulation, we can rule out __________- __________-__________-__________ if we find that the change is larger after the manipulation than the normal fluctuation observed during the pretreatment phase. __________ is not an issue because the same participants are followed throughout the study. Normally, one need not be concerned about __________ because the long pretreatment and posttreatment observation periods allow us to see naturally occurring changes or trends. However, two potential confounding factors are not easily controlled in a(n) __________- __________ design. These are history and instrumentation. History can be a confounding variable if any changes other than the manipulation occur at approximately the same time as the manipulation. In such a situation, it is impossible to know which factor accounted for the observed change in the __________ measure. __________ refers to the way the measurements are taken throughout the course of the study. Sometimes a manipulation is accompanied by changes in the __________ procedures, which can confound the results of the study.

  10. The interrupted __________-__________ design is flexible. It can be used to evaluate events that are actually __________ (such as a change in the law) or are uncontrolled (such as a natural disaster). It can be improved, however, by adding one or more __________ groups. These comparison groups may help to provide an indication of the effects of __________ variables that might influence the dependent measure. The comparison groups also provide some control for __________ in that they reveal general trends and variations in those trends over time.

  11. Program evaluation is one of the fastest growing areas of psychological research. The quality and __________ of a program should be evaluated so that budget dollars may be well spent. There are many practical problems associated with program evaluation research. __________ concerns are raised by the observations of individuals who have not consented to be part of a research project but, instead, are involved in a community program designed to meet some particular needs. A good program evaluator needs excellent __________ skills in order to gain the cooperation of the staff and clients of a particular program.

  12. Three principle ways of increasing the __________ in program evaluation are: (1) selecting appropriate dependent measures, (2) minimizing biases in the dependent measures, and (3) selecting an appropriate research design. It is best to select more than one __________ measure in order to capture the many aspects and effects of a program. Using __________ rather than subjective measures will tend to minimize normal biases in the dependent variable. Finally, using the strongest possible __________ design will give you the maximum control of extraneous variables and the greatest confidence in drawing causal conclusions.

  13. Many research designs have been used in program evaluation. The randomized __________- __________ design provides excellent control over extraneous variables. __________ considerations, however, will often dictate the nature of the control group. A single-group, __________-__________ design is a good alternative design. Here repeated measures of the dependent variable are made before and after the onset of the program. A pretest-posttest design, although commonly used, is not a very strong design and is not recommended.

  14. Survey research utilizes several basic procedures to obtain information from people in their natural environments. __________-__________ seek descriptive information about the current status of population characteristics. In contrast, __________-__________ tries to discover relationships among variables, as well as the current status of the variables.

  15. There are several steps in survey research. The first is to develop the survey __________. The exact questions and procedures that will be utilized to gather the information must be specified at this stage. The next step is to consider the __________ process. The type of sample will affect the conclusions that we can draw from the survey. The sample must adequately __________ the population to which we want to generalize. There are two types of sampling procedures. In __________-__________, we sample in an unsystematic "catch-as-you-can" manner and, consequently, cannot be sure that the sample is representative. In __________- __________, specific procedures are employed to assure that the sample is likely to be __________ of the population. This can be done with simple __________ sampling, in which every member of the population has an equal chance of being selected, or __________-__________ sampling, in which participants in the population are divided into subgroups and a random sample is taken from each of these subgroups. The larger the __________- __________ of a probability sample, the better the sample will represent the population.

  16. Several different research designs may be used with surveys. In the __________-__________ design the survey is administered once to a sample, yielding data on the characteristics of that population at that particular point in time. In the __________ design, also known as the "panel design," a group of participants are surveyed at several points in time. This design allows us to access changes within individuals over time.