• Charecteristics and Types
• Internal and External Validity
• True experimental designs
• Quasi experimental designs
• Evaluative and other designs
• References

EVALUATIVE AND OTHER DESIGNS

Evaluative designs

A. Overview of evaluative design

Evaluative studies are intended to provide data on the success of action programs.. These research designs are based on the objectives of the program and therefore several designs are based may be combined into one. Many clinical and educational programs have attempted to change something in the current system, improve some practice or enhance problem solving mechanism. These studies are both descriptive and experimental, testing the effect of some manipulation or change. Data may be either qualitative or quantitative and usually both are included.  To conduct evaluation research we need to be aware of the differences between using research principles for evaluative purposes and using them for projects that are strictly research endeavors.

B. Assumptions of the design:

• There are measurable objectives for the program that can be used as a basis for evaluation

• There are methods or tools available with which to measure the variables

• The objectives can be assigned priorities and weighed in a practical sense according to their value to the project.

• Adequate control participants can be provided so that a model for statistical testing can be used to establish whether or not the program made a difference.

C. Limitations

The ideal evaluative design has the major ingredients of an experimental design, but there is one more important difference. Because the study is evaluating a specific program, the program participants usually constitute the experimental participants rather than a sample of individuals chosen specifically for a research project. This is an important distinction because, the selection of these participants is usually determined by the program protocol, not the researcher. Thus they are not randomly assigned and there is no guarantee of equivalence between experimental and control participants.

Other designs

Kothari describes 4 formal experimental designs as follows

a. Completely randomized design:

Involves only two principles namely replication and randomization. The replications may be equal or unequal. It provides maximum number of degrees of freedom and is used when experimental areas happen to be homogeneous. There are two types. First one is two group randomized design in which sample is selected randomly, the experimental and control groups are given different treatments. The second is the random replication design in which a number of repetitions of the treatment.

b. The randomized block design

The randomized block design may offset the weakness in conventional randomization by grouping participants who share same characteristics so that like can be compared with like. Participants are matched and put into groups. The groups are then randomly selected to be either experimental or control. The number of groups depend upon, the purpose of the study. One of the limitations of this design is the difficulties involved in managing a large number of groups and in getting adequate sample sizes for each.

c. Latin squares design:

the treatments in this design are so allocated among the plots that no treatment occurs more than once in any one row or any one column. The two blocking factors may be represented through rows and columns. The merit is that it enables differences in a variable to be eliminated in comparison to the effects of different varieties of the same variable on the dependent variable. In this design one must assume that there is no interaction between treatments and blocking factors. The number in each treatment, row and column must be equal.

d. PRPP (Partially Randomized Patient Preference) design:

This design has the clear advantage in terms of persuading potential subjects to participate in a study because those with a strong preference get to choose their treatment condition. Those without strong preference are randomized but those with a preference are given the condition they prefer and are followed up as part of the study. The two randomized groups are part of the true experiment and the two groups who get their preference are part of quasi experiment. This type  of design can yield valuable information about the kind of people who prefer one condition over another but the evidence of effectiveness of the treatment is weak, because who elected a certain treatment likely differ from those who elected the other alternative.  These pre intervention differences rather than the alternative treatment could account for any observed difference in outcomes at the end of the study.

e. The Zelen design:

In the conventional RCT, participants who meet the inclusion criteria are randomized after they consent to take part. The process of seeking consent in this design depends on whether the single or double consent approach is used. In single version, participants in the control group are not asked consent, not made aware of the trial but included in the analysis. Only from those in experimental groups, consent is obtained. In the double version, all are asked consent, if some in experimental group refuse then they are offered the usual treatment/available alternatives.

Conclusion

controlled experiments are considered as the ideal methods in science. True experimental designs are the most powerful method of testing cause-and-effect hypothesis between variables. Although, experimental methods have many limitations. In health sciences, there are many constraints to do experiments and testing the results. Experiments are sometimes criticized for their artificiality.

References

1. Brink P J, Wood M J. Advanced design in nursing research. 2nd ed. New Delhi: Sage publications; 1998.

2. Parahoo K. Nursing research. 2nd ed. London: Palgrave macmillan; 2006.

3. Knapp T R.  Quantitative nursing research. New Delhi: Sage publications; 1998.

4. Grady K A, Wallston B S. Research in health care setting: applied social research methods series. Vol. 14. New Delhi: Sage publications; 1988.

5. Thomas B S. Nursing research. St. Louis: C V Mosby Co.; 1990.

6. Mateo M A, Kirchoff K T. Using and conducting nursing research in the clinical setting. 2nd ed. Philadelphia: W. B. Saunders CO.; 1999.

7. Senn S. Crossover trials in clinical research. New York: John Wiley and Sons; 1993.

8. Couchman W, Dawson J. Nursing and health care research, a practical guide. London: Bailliere Tindall; 1996.

9. Boniface D R. Experimental design and statistical methods. London: Chapman and Hall; 1995.

10. Woods N F, Catanzaro M. Nursing research theory and practice. St. Louis: C V Mosby Co.; 1988.

11. Lo Biondo-Wood  G, Haber J. Nursing research. 3rd ed. St. Louis: Mosby; 1990.

12. Notter L E, Hott J R. Essentials of nursing research. 4th ed. New York: Springer publishing co.; 1988.

13. Kothari C R. Research methodology. New Delhi: Wishwa prakashan; 1990.

14. Polit D, Beck C T. Nursing research. 8th ed. New Delhi: Lippincott Williams & wilkins; 2008.

15. Burns N, Grove S K. Understanding nursing research. 4th ed. St. Louis: Saunders; 2007.