Quantitative Research Article Critique
Susma Gurung
NUR 3130
New York City College Of Technology
04/15/2015
Quantitative Research Article Critique
Problem and Purpose
A problem statement is the expression of a problem and the opening point of a research. According to Polit and Beck (2014), it “articulates the problem and an argument that explains the need for a study.” (p.100). Therefore, a clear problem statement lets the readers know issues regarding the necessity of the particular research topic. In this article, “Psychological preparation for nasogastric feeding in children,” the researchers clearly stated the problem statement by explaining how there are only a small number of studies that have been done with regards to the psychological effects of nasogastric feeding in children (Holden, MacDonald, Ward, Ford, Patchell, Handy, Chell, Brown & Booth, 1997). In addition, such studies have often been found to be inappropriate to the level of cognitive development of the child.
In a quantitative study, the research goal, population of interest and the key study variables and their relationships are identified in the statement of purpose. The researchers clearly stated the purpose of the research by stating that “The aim of this study was to study the hypothesis that enterally fed children who receive detailed psychological preparation before feeding, and continuing supportive care subsequently, are less upset by the experience of nasogastric feeding than children who receive standard, low-intensity preparation and support.”(p.378). The population of interest was identified as children with nasogastric feeding. The key variables were children receiving detailed psychological preparation before feeding, and continuing supportive care subsequently (independent variable) are less upset by the experience of nasogastric feeding (the dependent variable). The verb ‘test’ suggests an experimental design.
The purpose of the research has significance for Nursing because it addresses the importance of psychological preparation for a vulnerable population during, what could be, a traumatic and invasive procedure. More nurses will be aware of the problem, and pathway for more research on this particular topic may be paved. Nurses can also use the results of this study appropriately as a guideline or reference when they have to deal with inserting a nasogastric tube for children. Armed with extra evidence based knowledge, nurses can be better prepared and make the enteral feeding experience for children less distressing.
Research design
This cross sectional, quantitative study used an experimental research design. The experimental group received detailed psychological preparation and support, while the control group received the standard low-intensity preparation and support. First, the children were placed in two different groups(Group A and B) according to their ages. After that, they were randomly allocated to the experimental and control group in their own groups. Therefore, every participant in Group A and Group B had an equal chance of being included in the experimental or control group. With randomization, Polit and Beck (2014) state that “there is no systematic bias in the groups with regard to attributes that may affect the dependent variable.”(p.154).
After the respective preparation and insertion of the nasogastric tube, blinded dietetic staff administered a detailed questionnaire to all the parents and older children in group B. Comparisons were made between Group A and Group B in order to enhance interpretability of the findings. The findings varied quite vastly across the two groups. In Group A, no statistical differences were found between the experimental and control group. A probable reasoning for that could be that the study had low statistical conclusion validity. However, in Group B, there was a statistically significant difference. One has to bear in mind that for group A, the only source of data collection was through the eyes of parents of the younger aged children.
The dissimilar results across Group A and B and the lack of statistical difference in Group A leads to the threat of internal validity. In this study, instrumentation threat is present. MacCoun (2010) states that in instrumentation, there are “changes in the measuring instrument (or the observer) that produce changes in the obtained measurements.” Questions like, how sure can the researchers be that a relationship really exists between the independent and dependent variable, come up. One way to improve instrumentation, would be to have the same qualified person observe and record findings. There was no mention of attrition/mortality. For external validity, the researchers mention that replication is needed so that the generalizability of the results would be increase. Since this is the first study that has been done to examine the role of preparation in children for nasogastric feeding, more research has to be done in order to improve external validity.
Population and sample
The population of a study is the entire group of concern while the sample is just the subset of the population (Polit and Beck, 2014). This study’s population was children requiring home enteral feeding which was well identified and described in detail. The sample was described in sufficient detail as well. The accessible population from which the sample was selected were from The Children’s Hospital, Birmingham NHS Trust. The sample of 48 children were divided into two groups, whereby group A consisted of 32 children, aged 2-6 years and group B consisted of 16 older children, between the ages of 7-16 years. Furthermore, the exact number of different gender of the children, the mean age of each group, and the indications for enteral nutrition was also stated. The study’s exact inclusion criteria was not clearly outlined but one can deduce that they were children aged between 2-16 years who needed home enteral feeding. On the other hand, the exclusion criteria, children who were already receiving nutritional support or had a psychiatric disorder history, was clearly stated.
A nonprobability consecutive sampling design was used in which the 48 consecutive eligible children were routinely allocated into the specified group according to their ages. Consecutive sampling is the stronger, better method of non probability sampling in controlling sampling bias because it allows for all the possible, pre-met criteria participants to be included. However, with consecutive sampling, it is important to note that variations may be seen during different seasons. Additionally, most times, nonprobability sampling hardly represents the population nor allows for everyone in the population to have an equal chance of being selected. Sampling bias is high which could result in misleading results. Therefore, utilizing a nonprobability sampling design is not the best possible sampling design when doing a research study. For the allocation of the participants into the experimental and control group, the researchers used random sampling, whereby every participant had an equal chance of being selected in each group. With random sampling, sample biases would be minimized because any differences that are seen in the sample and population can be proclaimed as pure coincidence. Furthermore, the extent of sampling error can be approximated.
A power analysis is a critical and essential part of the sample size in quantitative studies and it is normally utilized by researchers to estimate how big their sample size ought to be when testing their research hypothesis. In this study, a small sample of 48 children was used, but there was no mention of the usage of a power analysis to estimate the sample size needs. The lack of a power analysis might yield results that are not statically significant. However, a pilot study was done at The Children’s Hospital, Birmingham NHS Trust from which the researchers got conflicting views regarding the need for psychological preparation for nasogastric feeding. Even though the study did not mention a power analysis, using a pilot study, the researchers might have been able to estimate how large the group difference would be, which is key in using power analysis. Polit and Beck (2014) mentioned that with such a small sample being used, “the risk of “getting it wrong” (statistical conclusion validity) increases….” (p.182). I believe that the sample size was not adequate because of the lack of clarity and rationale in justifying the sample size and sample selection method. With a small sample size or weak sampling strategy, biasness might occur. The researchers themselves explicitly stated that the results of the study has to be taken into account carefully due to the small sample size.
Data collection and measurement
Data collection is an important aspect of a study and there is a variety of data collection methods that can be adopted to collect data. In this article, “Psychological preparation for nasogastric feeding in children,” key variables were operationalized through observation and a questionnaire, a self report data which is one of the most common data collection approach in nursing studies. The questionnaire incorporated various types of data collection methods like the Likert scale, open-ended questions, a priority rating scale and play performance scale. The researchers adequately described what each section comprised of. In section 1 of the questionnaire, the Likert scale with 5 declarative statements was used where respondents had to choose how much they agree or disagree with that declarative statement. Section 2 included open-ended questions which might be more difficult to analyze given the varying answers. Additionally, a researcher’s bias might affect how the responses are deciphered. However, more extensive information may be yielded using this open-ended approach.
Section 3 was an example of a closed-ended question instrument whereby respondents were asked priority and rating questions. Such close ended questions ensure response comparability and aids analysis (Polit and Beck, 2014, p. 184). Moreover, they are more time efficient. However, a down-fall is that some crucial responses might be excluded. The Lansky play performance scale is an example of direct observation of behavior using a rating scale. I believe that the use of structured observation was a good alternative to self reports because some of the children would not be able to communicate as needed due to their age. Scale devices allow researchers to quantify responses much more easily but can result in response set biases as well.
Data collection was only done after the enteral feeding experience. There was triangulation of method which is desirable. The multiple methods were used appropriately because each method gave a different view on the effects of the enteral feeding experience based on the different preparation that one received and the age. A self report questionnaire can be used to gather information through relevant question. However, a crucial weakness of the self report questionnaire data is the validity and accuracy. Respondents can falsify answers and social desirability bias can occur (West, 2014). There is almost no way of finding out how trustable and honest the answers given in the study were. Although there was an appropriate mix of questions in the structured self reports, there was no mention of any efforts from the researchers to enhance data quality like looking at ways to decrease response biases. Additionally, there was no mention of the data collection methods yielding data that were high on reliability and validity.
Procedures
The intervention in this study was the children in the experimental group receiving detailed psychological preparation before feeding, and continuing supportive care subsequently. The intervention was adequately described and implemented separately for the two different age groups. The researchers stated that the child’s age and developmental status was addressed in detail. However, a key detail of this study is the lack of a discussion about intervention fidelity. It is crucial to address intervention fidelity so that readers can judge the quality of the study and aid researchers in understanding how different factors may have played a role in the outcome of their study (Gutam & Murphy, 2012). In addition, biasness can be averted. The researchers utilized a play therapist and nutritional care sister for both groups but the experience and training of the professionals were not mentioned. Nevertheless, a detailed explanation of the intervention was explained which can ensure consistency of intervention delivery. All the participants in the experimental group received the interventions that were mentioned.
As mentioned earlier, data collection was done through observation and a questionnaire. Blinded dietetic staff administered the questionnaire in order to minimize bias. However, once again, a flaw in the study is that there was no mention of the amount of training the staff had. With the play performance scale, the observers were parents which could have its pros and cons, in the sense that a parent is most likely the best candidate to evaluate their child’s behavior. However, with that same mindset, with the amount of emotions and personal values involved, it might cause the parent to be vulnerable to biasness. There was a brief description of what the parent should be observing in the child’s behavior but observational bias was not addressed at all.
Data Analysis
Statistical analysis is a vital part of every research paper. According to Frost (2013), ” understanding statistics can help you assess the quality of other studies and the validity of their conclusions.” In this study, statistical analysis for the questionnaire of both groups was done using different methods. For group A, the chi-squared test was used while for Group B, the median test was used. According to Polit and Beck (2014), the chi-squared test “is used to test hypotheses about the proportion of cases in different categories..” (p.235). The exact chi scores were not given for either groups. For the questionnaire data in group A, there were no statistical differences between the experimental group or the control group. One reason for that could be the small sample size and the lack of a power analysis. However, for group B, using the median test, the researchers stated that there was a statistically significant difference whereby the experimental group who had received detailed preparation were significantly less distressed. The Mann-Whitney U test was used to analyze Lansky play scale scores where both group A and B had no statistical differences between the experimental group or the control group. Due to the different tests being used to analyze the questionnaire data, one can only wonder if that might have played a factor in the differences of data analysis between the two group. It is ambiguous if the analysis controlled for confounding variables. The researchers explicitly stated that the results of the study has to be taken into account carefully due to the small sample size and the fact that routine preparation and psychological support appeared to be beneficial for the older children. It was not clearly stated whether or not the hypothesis was supported but one can presume from the data analysis that there was mixed results in which the hypothesis was supported only for older children.
Findings
Some information about the statistical significance was presented in the study. As mentioned earlier, in Group B, there was a statistically significant difference whereby the experimental group who had received detailed preparation were significantly less distressed. The p value is stated as less than 0.05 which means that there is a less than 5 in 100 chance of a bogus result. Therefore, we can conclude that the study’s findings are probably true and can be replicated with a new sample (Polit & Beck, 2014).
The results were mixed and the researchers mentioned that the younger children’s feelings and reactions were based only on the parent’s words. The fact that the questionnaire was filled out by different people, parents in group A and the older children themselves in group B, could therefore account for the mixed results. There is no mention of the precision of estimates.
Summary assessment
I believe that the study findings does not really appear to be very valid or have much true values. I say that in part because of the small sample size of the study and lack of statistical difference across the different age groups. The age difference might have contributed to the fact that different respondents answered the questionnaire and that different tests were used to analyze the questionnaire data. With that being said, it becomes unclear if those might have played a factor in the varying results.
Regardless, I believe that this study has an important effect in nursing practice. The evidence might not be very strong but light is being shined on the importance of psychological preparation for a vulnerable population during, what could be, a traumatic and invasive procedure. Therefore, it might motivate others to conduct more time and research on this matter. Nurses can also use this study aptly when having to insert a nasogastric tube for children.
References
Frost, J. (2013, March 4). Why Statistics Is Important. Retrieved May 10, 2015, from http://www.worldofstatistics.org/2013/03/04/why-statistics-is-important/
Holden, C., MacDonald, A., Ward, M., Ford, K., Patchell, C., Handy, D., . . . Booth, I. (1997). Psychological preparation for nasogastric feeding in children. Children’s Nursing, 6(7), 376-385.
MacCoun, R. (2010, November 5). Experimental Design Strategies. Retrieved April 05, 2015, from https://www.law.berkeley.edu/files/MacCounNov5_slides_handout.pdf
Murphy, S. L., & Gutman, S. A. (2012, July 1). Intervention Fidelity: A Necessary Aspect of Intervention Effectiveness Studies. American Journal of Occupational Therapy, 66, 387–388. doi:10.5014/ajot.2010.005405
Polit, D., & Beck, C. (2014). Essentials of Nursing Research: Appraising Evidence for Nursing Practice. Philadelphia. Wolters Kluwer Lippincott Williams & Wilkins.
West, M. R. (2014, December 18). The Limitations of Self-Report Measures of Non-cognitive Skills. Retrieved May 2, 2015, from http://www.brookings.edu/research/papers/2014/12/18-chalkboard-non-cognitive-west
