Article
The first article to be criticized is titled “Should we lower the dose of iron when treating anemia in pregnancy? A randomized dose-response trial” by Zhou, S.J Crowther, C.A, Gibson, R.A and Makrides, M published in 2009.
Sampling Inclusion Criteria and Exclusion Criteria
The inclusion of pregnant women who were attending antenatal care diagnosed with anemia being included randomly is appropriate for the study. Similarly, the process of excluding those suffering from the disease but iron deficiency not being the cause has previously abused alcohol, is diabetic, and if they were taking supplements vitamins as well as mineral ones. In my view sampling has been identified
Sample Size
For this article, the authors did identify the sample size, the women selected were 180. However, the response rate was 99.0% as only 179 were able to complete the trial. It is worth noting that the authors’ research designed included dividing the sample selected into three different groups based on the concentration of ferrous sulfate which was 20, 40, and 80mg per tablet. These three distinct treatments or groups have been thoroughly explained by the authors making it possible for other researchers to replicate the methodology used.
Power analysis reported
Power analysis a technique used in experimental research aimed at estimating the desired sample size has not been reported for this article. From my judgment, it is evident that the sample size used although there was no power analysis is appropriate since the findings have been consistent with previous findings.
Attrition
Just like most research studies, there are chances of the selected respondent to refuse to respond, die, or move out of the study are. For this study, the response rate was high, 99.0% as the only 0ne out of 280 who gave their consent moved out of the study are. She sought antenatal care and birth in another hospital.
Characteristics of the sample
The authors of the article did a good job of giving a detailed demographic characteristic of the respondent in the three distinct categories or groups. It is worth noting that such an arrangement has made it easier for one to easily compare among the groups. The summary of all this is provided in Table 1 Trial entry characteristics of participating women.
Sampling method and adequacy
Sampling which is the selection of a representative of subjects to be used in research is of paramount significance. For this study, the selection of only women being diagnosed with anemia at the Children, Youth, and Women Health Service and the mid of their pregnancy seems to be non-probability sampling. The specific type of non-probability technique used is purposive sampling methods which are usually carried out with some purpose in mind (Zhou et al., 2009). Based on the type of sampling methods, it is very challenging to establish whether the sample selected is a representative of the sample. The best techniques to certain this could be random sampling.
Sample Representative of the Accessible and Target Populations
As stated previously, the non-probability sampling technique used in this experimental research leaves very little room for one to establish whether or not the sample is representative of the general population. However, there is room for the sample to be representative of the accessible and target populations although this is not guaranteed like in random or probability sampling technique.
Possibility of Type II error
There are two main types of errors ion statistics, type I, and type II. In this article, the sample size chosen can result in a type II error. It is worth noting that, when the sample size is too small there is a higher risk of committing type II error usually resulting in failing to reject a null hypothesis when in fact it opts to be rejected. Since the sample size was only 180, the chances of the error are high (Zhou et al., 2009).
Potential Biases in the Sample
There are no studies that do not face biasness. For this article, although not clearly stated, one can identify several biases brought forth by the authors indirectly. For instance, the definition of anemia seems to have generated biasness. Additionally, self-reporting of such events as the possibility of side effects after assessment and the sample size are some of the contributions of biases although not succinctly discussed by the authors.
The Target Population
For every research study, be it experimental, nonexperimental, survey, quasi-experimental, there is always a target population that the finding is generalized (Polit & Beck, 2007). The authors of this article have made it clear that the target population is medical service providers as well as pregnant women and those who intend to procreate.
Article 3
The second article being criticized is entitled “The Effects Of Weekly Iron And Vitamin A Supplementation On The Hemoglobin Levels And Iron Status In The Adolescent Schoolgirls In Western Kenya”. By Leenstra, T., Kariuki, SK, Kurtis, JD, Oloo, AJ., Kager, PA., & Kuile, FO.
Sampling Inclusion Criteria and Exclusion Criteria
The inclusion of school girls aged 12-18 in a randomized manner was appropriate for the study. Similarly, the process of excluding schools with more than 10 girls fulfilling the entry criteria was also much significant due to logical reasons that the study was to incur. Also, the other most important thing that the researcher did was to exclude girls experiencing severe VAD, pregnancy, or even concomitant diseases which required hospitalization (Leenstra, et al, 2009). Due to these exclusions and inclusions, I view the sampling well identified. In my view sampling has been identified.
Sample Size
The researchers of this article, in a clear manner, identified the sample size they were going to use. They used an estimation of 220 participants, with an average of 55 per arm in the detection of a 0.5gl-1. The anticipated drop out was placed at 20%. It is at the same time much significant mentioning that, the researchers assumed a standard deviation of Hb means level at 0.8gdl-1.
Power Analysis Report
Power analysis is the technique that the researchers used in the estimation of sample size, it is very important that the researchers estimated it at 90 percent and 95 percent confidence. To my view, a power analysis was important in ensuring that, the findings are consistent with previous findings.
Attrition
Since there were chances that some girls will pull off themselves from the study, as it has been evidenced in previous studies, the researchers estimated some changes from the mean. As a result, the anticipated drop out was placed at 20%. They also assumed a standard deviation of the average Hb level at 0.8gdl-1
Sample Characteristics
It is worth mentioning that, the researchers did a good job of explaining the characteristics of the respondent. They mentioned that the respondents were school girls aged between 12-18 years, in fourteen primary schools. As a result, all the groups the arms that the researchers mentioned in the study design and recruitment had comparable traits.
Sampling Methods and Adequacy
As defined earlier sampling is just a method of selecting subjects to be used in the study, hence is of great importance. In the study, the selection of schoolgirls, the exclusion of schools with more than 10 girls showing entry criteria, as well as the exclusion of severe VAD, depict that the method is non-probability. The best method used in the study to adequately obtain a representative was random sampling.
Sample Representative
Since the study used a non-probability sampling technique, the experimental research has not identified if the sample is representative of the target population.
Possibility of Type II Error
In this study, there are some chances of committing a type II error. Since the sample was 220, which to my view is very small, there are chances of committing type II error.
Potential Biases in the Study
Though not yet outlined in the article, generally most studies face the problem of baseness.
Target Population
Those who carried out this study clearly defined the target population as being school girls aged 12-18, from schools having less than 10 girls with fulfilling entry criteria. They also added that they should have no severe VAD, pregnancy, or concomitant diseases that require hospitalization.
4th Article
The fourth article to be criticized is called The Overlap Of Overweight And Anemia Among Women In Three Countries Undergoing The Nutrition Transition, by Eckhardt, CL., Torheim, LE., Monterrrubio, E., Barquera, S., &Ruel, MT.
Sampling Inclusion and Exclusion Criterion
The inclusion of only non-pregnant adult women aged 18-49 was very significant, as they hand information on all variables that were of interest. Though the exclusion of underweight women from regression was of importance, it was not of paramount importance in the study. As a result, both exclusion and inclusion criteria have been identified.
Sample Size
The researchers who carried out this study were identified. They clearly stated sample sizes for every group that was selected for this study. For instance, in Mexico, the sample size was 11965, In Egypt, the sample size was 6841, and the sample size in Peru was 5078 women. The three demographic groups have been clearly explained.
Power Analysis Report
The authors have this article explain any power analysis report, hence leaving very little room for the establishment if the sample size was appropriate based on power analysis. However, since the results obtained were consistent with previous studies, then the sample size was appropriate for this study.
Attrition
Though there were chances of the selected population pulling out due to their reasons, the authors did not discuss the refusal rate and sample mortality.
Sample Characteristics
The authors of this article moved a step higher in the explanation of demographic characteristics of the sample in three different countries, which has enhanced easier comparisons among the categories. According to “table 2, the prevalence of Anemia” (Eckhardt, et al 2008), the three demographic groups had comparable characteristics.
Sampling Methods and Adequacy
Due to the importance of sampling technique in any research, this study has explained clearly that, probability sampling method has been used. It explained that a probabilistic sample of about 21503.
Sample Representative of the Accessible and Target Population
As stated earlier, the probability sampling technique used in the research, clearly shows that the sample is representative of an accessible and target population.
Possibility of Type II Error
It is good to note that, when the sample size used is too small, then there are high chances of encountering type II error, however, since the sample used in this article is big enough, the probability of committing type II error is too small. This is based on the fact that the sample size used in Mexico was 11965, 6841 in Egypt, and 5078 in Peru. Such samples are large enough to reduce the type II error.
Potential Biasness
Though not clearly explained in this article, just like any other studies, there is indirect biasness. This is for instance experienced when defining differentiating urban and rural areas in the three regions.
The Target Population
Those who carried out the research presented in this article started that in every household, they targeted data collection from all children from 0 to 11 years, but only from one woman aged 12-49 years, who had information concerning all variables of interest. However, for analysis reasons, the target population included non-pregnant women aged 18 to 49 years. The entire target population was from three countries undergoing the nutrition transition.
References
Eckhardt, C. et al. (2008). “The Overlaap Of Overweight And Anemia Among Women In Three Countries Undergoing The Nutrition Transition”. European journal of clinical nutrition. 62, 238-246.
Leenstra, T. et al. (2009). “The Effects Of Weekly Iron And Vitamin A Supplementation On The Hemoglobin Levels And Iron Status In The Adolescent Schoolgirls In Western Kenya”. European journal of clinical nutrition. 63, 173-182.
Polit, F. , Beck,C. (2007). Essentials of Nursing Research. 6th edition. Baltimore: Lippincott.
Zhou, S.J. et al. (2009).“Should we lower the dose of iron when treating anemia in pregnancy? A randomized dose-response trial”. European Journal of Clinical Nutrition, 63(1): 183-190.