Marketing Research III
188 PEDIATRIC NURSING/July-August 2011/Vol. 37/No. 4
D emystifying the research process often involves understanding research terminology, the rationale for the selection of a research design, and the known benefits and consequences in the selection
of a design. This commentary discusses the major aspects of a well-known and used quantitative research design in nursing research used by Tourigny, Clendinneng, Chartrand, and Gaboury (2011) to evaluate the utility of a virtual education tool for pediatric patients undergoing same-day surgery. The rationale for why this design was chosen by these nurse researchers and its advantages and disadvantages are discussed.
A research design is the overall plan for answering research questions and hypotheses. The design spells out strategies the researcher adopts to gather accurate, objec- tive, and interpretable information (Polit & Beck, 2007). Tourigny et al. (2011) used a non-experimental, quantita- tive research design known as a descriptive, comparative design. It is also known as casual comparative research and pre-experimental research. The basic purpose of these designs is to determine the relationship among variables. The most important distinctions between these designs and experimental designs are no control (manipulation) of the independent variable (IV) and no random assignment of study subjects to the intervention or control group. These designs are frequently used in nursing research studies because nurse researchers are often faced with these specif- ic limitations.
In summary, the known properties of descriptive, com- parative research studies are 1) no manipulation of an inde- pendent variable, 2) no random assignment to groups, and 3) often inclusion of a control or comparison group. The paradigm for these studies is diagrammed in Figure 1.
In this diagram (see Figure 1), the researcher hypothe- sizes that “X” is related to and a determinant (cause) of “Y,” but the presumed causes are not manipulated, and subjects are not randomly assigned to groups (LoBiondo-Wood & Haber, 2010). Rather, a group of subjects who has experi- enced “X” in a natural situation is located, and a control
group of subjects who has not experienced “X” is chosen. The behavior performance or condition of the two groups is compared to determine whether the exposure to “X” had an effect predicted by the hypothesis (LoBiondo-Wood & Haber, 2010). Tourigny et al. (2011) hypothesized a deter- minant of study participants’ level of knowledge about hos- pital equipment and procedures, and their emotional state would differ based upon whether or not they viewed the Surgery Virtual Tour presentation. In this study, the exposed group resulted from participants choosing to view the Surgery Virtual Tour. These researchers then compared this group with a group at the same institution who did not view the Surgery Virtual Tour presentation.
Tourigny et al. (2011) noted that the Surgery Virtual Tour was posted on the hospital’s Web site and available to all children, adolescents, and parents being cared for at this institution. Thus, these researchers had no control over which study participants viewed or did not view the educa- tional program. Prohibiting access of this educational pro- gram to some participants for the purposes of conducting this research study would have violated these children’s, adolescents’, and parents’ ethical right to fair treatment. The right to fair treatment is based on the ethical principle of justice that each person should be treated fairly and should receive what he or she is due or owed (Burns & Grove, 2005).
An important criterion in determining a research design’s rigor is its potential to generate findings that are interpretable. The term interpretable relates to the credibil- ity and dependability of data generated by a study, and is based on the study’s design to sufficiently test “cause and effect” relationships. The term “causality” implies that a systematic relationship exists between the independent variable (IV), which is the “cause” or intervention of the study design, and the dependent variable(s) (DV) or the outcome(s) of the study. In other words, confidence that the outcome of a research study is a consequence of the effects of the intervention must exist.
Mary Ann Cantrell, PhD, RN, is an Associate Professor, College of Nursing, Villanova University, Villanova, PA.
Demystifying the Research Process: Understanding a Descriptive Comparative
Research Design Mary Ann Cantrell
Demystifying Research
Tourigny, J., Clendinneng, D., Chartrand,C., & Gaboury, I. (2011). Evaluation of a virtual tool for children undergo- ing same-day surgery and their parents. Pediatric Nursing, 37(4), 177-183.
With Demystifying Research, nursing research leaders comment on some aspect of a research article featured in the issue, with the aim of helping the reader better understand research. Look for Demystifying Research in each issue of Pediatric Nursing.
PEDIATRIC NURSING/July-August 2011/Vol. 37/No. 4 189
There are three criteria for causality: 1) the cause (the IV) must precede the effect (the DV) in time, such that the IV had to occur before the DV); 2) an empirical relationship exists between the IV and DV, meaning that a relationship that is measurable must exist between the presumed cause and effect; and 3) the relationship between the IV and DV cannot be explained by a third variable. Of these three cri- teria, researchers are most concerned about ensuring results of their study are due to the experimental treatment and not due to the characteristics of the subjects or other com- peting explanations for the results. Characteristics of the subjects or other competing explanations are known as internal validity threats.
There are several limitations in the design used by Tourigny et al. (2011) that threaten the confidence in their study’s findings, specifically having no control over the internal validity and characteristics of the subjects influenc- ing the outcome of the study. The internal validity threat due to characteristics of the subjects is known as selection bias and is always a threat if random assignment to groups does not occur. Researchers are cautioned to be aware that when intact groups are compared, differences existing between the two groups before the start of the experiment could have affected the outcome of the study. People “self- select” to a group based on personal characteristics and preferences, and these personal characteristics and prefer- ences can influence the outcome of a study. Tourigny et al. (2011) addressed this potential threat operating in their study’s findings by measuring selected differences in socio- demographic variables that could have accounted for dis- similarities between the groups. There were no significant differences in socio-demographic variables between partici- pants who viewed or did not view the Virtual Tour, with the exception that families who took the Tour were more like- ly to have access to the Internet at home (Tourigny et al., 2011). These findings provide some evidence that these socio-demographic variables can be ruled out as internal validity threats operating in this study; however, it remains unknown if characteristics not measured by Tourigny and colleagues could be operating as threats to the study’s inter- al validity. It is not feasible to measure an exhaustive list of socio-demographic characteristics that could pose every possible internal validity threat related to study partici- pants’ characteristics, but researchers carefully select known factors from previous studies and their clinical experiences as was done by Tourigny et al.
Another strategy used by Tourigny et al. (2011) to increase the internal validity of their study was to establish inclusion and exclusion criteria to determine the study’s sample. Inclusion and exclusion criteria are guidelines or the standards determining who can or cannot be in the study. Population descriptors, also known as important characteristics of a population, are criteria that set the stan- dards. These characteristics can also operate as internal validity threats in a study. In their study, Tourigny et al. identified the inclusion criteria for their study as only allowing children and adolescents 6 to 18 years of age, able to understand or read and write in English, be at a school-
age cognitive level, and who gave an assent or written con- sent to be in the study. They also excluded children with any developmental or physical state that could prevent them from completing the questionnaires. These criteria placed more control over potential internal validity threats operating in the study, but as a consequence of doing so, the external validity of the study’s findings was decreased. External validity addresses the ability to generalize the find- ings of the study to other groups. The findings generated by Tourigny et al. are not generalizable to children younger than 6 years, who are unable to understand or read and write in English, are not at a school-age cognitive level, or have a developmental or physical impairment. Internal and external validity have an inverse relationship; the more internal validity control a study design employs, the more likely its external validity will be limited.
In summary, Tourigny and colleagues (2011) selected a feasible research design; its implementation protected research participants’ ethical rights, tested the identified intervention, and generated interpretable findings. A researcher’s choice in selecting a research design is depend- ent on many factors, and researchers usually make con- scious decisions in their selection to augment some aspects of rigor in their study while foregoing others. Selection of a research design requires creativity to maximize inter- pretable findings within known limitations in conducting the investigation.
References Burns, S., & Groves, S.K. (2004). Understanding nursing research (3rd
ed.). Philadelphia: Saunders. LoBiondo-Wood, G., & Haber, J. (2010). Nursing research: Methods
and critical appraisal for evidence-based practice (7th ed.). St. Louis, MO: Elsevier.
Polit, D.F., & Beck, C.T. (2007). Nursing research: Generating and assessing evidence for nursing practice (8th ed.). Philadelphia: Lippincott Williams & Wilkins.
Tourigny, J., Clendinneng, D., Chartrand, C., & Gaboury, I. (2011). Evaluation of a virtual tool for children undergoing same-day surgery and their parents. Pediatric Nursing, 37(4), 177-183.
Figure 1. The Paradigm for Descriptive, Comparitive Research Studies
Groups X (The “Cause)
(The “Intervention”) Y (The Effect)
(Outcome Variable)
Exposed group Exposure to “X” Measurement of “Y”
Control group No exposure to “X” Measurement of “Y”
Demystifying the Research Process: Understanding a Descriptive Comparative Research Design
Pediatric Nursing Call for Evidence-Based, Clinical
Manuscripts for These and Other Topics
• Allergies • Alternative and
complementary treatments • Burns assessment • Calcium and school aged girls • Care of the child with a
congenital heart defect • Care of children with
neurological problems • Care of the child with diabetes
mellitus
• Celiac disease identification and long-term ramifications
• Chronic illness/nursing management
• Clinic aspects of care in ambulatory setting
• Dealing with difficult family • Discharge of the complex
patient • Diseases/disorders associated
with pain, such as osteogenesis imperfecta
Questions regarding these or other topics can be directed to Judy A. Rollins, PhD, RN, Editor of Pediatric Nursing (rollinsj@msn.com). A copy of the author guidelines can be obtained from the journal Web site (http://www.pediatric nursing.net), or will be forwarded at your request by contacting the journal office at Pediatric Nursing, Jannetti Publications, Inc., East Holly Avenue/Box 56, Pitman, NJ 08071-0056. Phone: 856-256-2300. E-mail: pnjrnl@ajj.com. Pediatric Nursing is a refereed, bi-monthly publication.
Copyright of Pediatric Nursing is the property of Jannetti Publications, Inc. and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.