1. In your own words explain what a good experimental design should accomplish.
A good experimental design needs to divide the participants into two main groups, the control group, and the experiment group. The experimental design should introduce a change into the experiment group rather than the control group. It should give estimates that are not biased of the uncertainties or factor effects. For natural interpretation and understandability, the experiment should be simple and present results that are easily interpretable. If the differences are noticed between the two participants group, the research needs to detect these differences.
2 . Describe some advantages of experimental designs that combine analytic tactics. 229-230
Delegate your assignment to our experts and they will do the rest.
Multiple baselines across interventions are essential as it helps conduct an experiment when the research question is too broad to be accommodated in a single experiment design. It is also useful when the outlined procedures to be used are too complicated. The combined analytical process helps solve the problem of a very complex system that cannot accommodate a single design. It enables the experiment to be conducted conclusively to give a convincing result.
3 . Be able to cite one study (mentioned in the book) that used a design combining experimental tactics; be able to cite the purpose of the study, the dependent measure(s), and the experimental design used. 229-230
Sisson and Barrett (1984) incorporated multiple probes across behaviors components and alternative treatment analysis, and a numerous baselines across behaviors elements in a design comparing the effects of two language training procedures. The experiment permitted the experimenter to identify the superiority of the total communication methods for the two children. It also shows that direct application of the treatment was necessary for training to occur on specific sentence structures.
4. Describe the concern behavior analysts (and others) have about the traditional group designs. 227-228
Behavior analysts are concerned with improving the behavior of individuals. They know that an average change in the group performance may not show any modification of the individual subjects. The average may change, but it cannot reveal what exactly happened to the individual subjects whose performance may either increase or decrease.
5. Describe some disadvantages of using statistics in analyses. 227-228
Statistical analysis is unable to correct the variability witnessed in the individual's performance in a group experiment. It cannot remove the influence already represented in the data (Johnston and Pennypacker, 1993, p. 184). Statistical analysis cannot cancel out the order effect in a group experiment design. Even after averaging the data together statistically, the order effect remains unless one variable is held constant. Also, the use of statistical analysis leads to loss of the power of replicating effects with individual subjects. Replicating impact is one of the most strength for a convincing demonstration.
6. Restate the definition of internal validity.
Internal validity is the extent to which one can approximately conclude that a change in one variable X, has led to a change in another variable Y. Variable X can be manipulated and observe the change in the variable Y. The degree of this change is the internal validity.
7. State the four elements of an experiment that could be related to internal validity issues.
The measurement is taken during the experiment.
The subject of the study.
The setting of the experiment.
Independent variable.
8. For each of the four elements related to internal validity, define, give examples of potential internal validity threats associated with this aspect, and describe some ways of minimizing the influence of this component on the outcome of an experiment.
Measurement is the amount or quantity of each variable used in an experiment. Examples of internal validity threats in measurement include observer drift and observer biases. This threat can be solved by the use of intermittent probes by the observer. A random sample can be used in eliminating bias.
The subject is the primary element under study. It is what is being studied. Change in the subject over the course of study can create a threat to internal validity. However, the correction of this requires using one subject throughout the experiment.
The setting of the experiment is the environment in which the test is taking place. Studies conducted in natural environments are more likely to cause threats to internal validity. It can be solved by conducting the studies in a controlled area such as the laboratory.
Independent variable is a variable that does not depend on another variable for it to change. It is always fixed. When these variables are complicated, they can be a threat to internal validity. It can be solved by having some control over the independent variables.
9. Define double-blind control, treatment integrity, treatment drift, and procedural reliability. Give examples of each.
Double-blind control is an experiment in which the possibility of biases is elimination by keeping both the subject and the experimenter unaware of what is being tested. An example is a medical test experiment.
Treatment integrity is the extent to which independent variable is implemented as required.
Treatment drift is a situation where the application of independent variable during later phases of an experiment is different from the way it was applied at the start of the study.
Procedural reliability is the consistency with which research has been conducted.
10. Define external validity and describe how it is different from internal validity. If given an example, be able to accurately label it as illustrating either internal or external validity.
External validity is the extent to which the outcome of an experiment can be generalized to other situations. The main difference between internal and external validity is that internal validity deals with the relationship between dependent and independent variables while external validity is concerned with making inferences about the whole population and universality of the outcome.
11. Define, describe, and contrast the different types of replication. Give examples of each type.
Replication is the process of repeating the experiment using the same methods and techniques but with different subjects and experimenters. The same test is repeated with various subjects and experimenters to be able to affirm that the results obtained are valid and accurate. The main types of replication are direct and systematic replication. Indirect replication, the research duplicates precisely the previous study with either same subject or a different subject. The example involves a survey of the language training among children. In systematic replication, one aspect is purposefully varied.
12. Define what is meant by “social validity” and give examples of results that might be or might not be socially valid. Describe how social validity is measured, and describe the point of the three topics under this heading.
Social validity is the change in a person’s life in a direction that is helpful to the person. Examples, intervention for children’s behavior are socially valid.
After developing the effective behavior intervention plans, the social validity of the interventions is compared using the three measures namely concurrent selections, verbal reports, and maintenance of the intervention over a given period.
The social importance of the goal, social appropriateness of the process and procedures and the social significance of the effects are also considered.
13. Min. 300 words for this question (1 page)
Briefly define the following, internal validity, external validity, and social validity. Then explain how each form relates to and differs from the others. Give a specific example of how a study could have strong internal validity and little to now social validity or how a study could have great social validity and have poor external and or internal validity.
Internal validity is the extent to which an experiment shows convincingly that changes in a variable are a function of the independent variable and the result of an uncontrolled or unknown variable. External validity is the extent to which the outcome of an experiment can be generalized to other situations. The main difference between internal and external validity is that internal validity deals with the relationship between dependent and independent variables while external validity is concerned with making inferences about the whole population and universality of the outcome. Social validity is the change in a person’s life in a direction that is helpful to the person.
The study needs to have a great internal validity by having control over various threats. Biases during measurement can lower internal validity. A random sample needs to be taken to overcome the biases threat to internal validity. Having some control over the independent variable can also help create a high internal validity. The experimenter needs to establish some control over the independent variable. Strong internal efficacy can be achieved by having a normal subject throughout the study. Changing the subject will lead to weakening the internal validity. The environment in which the research is taking place can also influence the degree of internal validity. A study conducted in a natural environment is more likely to have a weak internal validity. The study needs to be done in a controlled setting such as the laboratory to give a high validity.
Example, in the study to identify the relationship between the learning rate and the method of teaching in the classroom, the learning rate becomes the dependent variable while teaching method the independent variable. This study should be conducted by taking a sample of students and interviewing them. The sample will eliminate bias. The interview should be done in a controlled environment to enable the students to give their opinion freely. The subject should remain the same throughout the study period.
