The Paper was written by Li et al. (2020) focuses on the study of bipolar disorder as one of the most common diseases associated with the human psyche. However, despite its prevalence, this disorder is often misdiagnosed, or no signs of the disease are noticed at all. Thus, the researchers’ goal is to develop a support vector machine (SVM) model that would help to accurately identify patients with this condition (Li et al., 2020). It uses a combination of structural and functional MRI based on machine learning technology.
Several methods and techniques have been used to achieve this goal. The study involved 80 patients, of whom 44 had bipolar disorder, and 36 were in the control group. For each of them, clinical studies were carried out combined with MRI scans (Li et al., 2020). Based on the data analyses obtained with the help of VBM and ReHo, two t-tests were carried out, which made it possible to select the final data clusters necessary for the formation of the SVM model.
The model compiled from this study identified patients with bipolar disorder with extremely high efficiency. According to the results, the identification accuracy was 87.5%, the detection sensitivity was 86.4%, and the specificity was 88.9% (Li et al., 2020). Therefore, the authors successfully proved that a combination of structural and functional MRI could improve the accuracy of determining bipolar disorder in patients. Thus, the task set by the authors was completed, the results were processed, and the information obtained can be used to implement further programs for the study of bipolar disorder.
As mentioned above, in this study, a two-sample t-test with multiple comparisons was used as a statistical analysis method. This approach is well suited to the existing conditions due to the nature of the study. The T-test is used to analyze the differences between the means of the two study groups. Since the study involved a group of people with bipolar disorder and a control group, the use of the t-test seems to be the most acceptable. In addition, the sample sizes are small enough to allow detailed analysis of the distribution.
On the other hand, this method can be replaced by other forms of statistical analysis. First of all, in this context, a z-test can be used, which is closely related to the used t-test. However, to implement this method, researchers would have to make adjustments to the research. It would be necessary to increase sample sizes since the z-test works well on large enough groups. In addition, it is assumed that when using this technique, the standard distribution of the population is known, which is absent in the current version. Finally, as the study expands further and the number of variables and groups increases, the variance or ANOVA test analysis can be used.
A study by Saddik et al. (2021) focuses on the effects of the raging COVID-19 pandemic on the psychology of adults and children in the United Arab Emirates community. According to the researchers themselves, although the psychological impact of other similar infectious diseases is well studied, there is very little information regarding COVID-19 in the UAE. Accordingly, current research does not assess the potential risks that this pandemic may pose to the community. Consequently, the work aims to research the level of anxiety among children and adults living in the UAE.
To achieve this goal and obtain data for analysis, the article’s authors conducted a cross-sectional web survey. The subjects of this survey were 2,200 adult volunteers and their children. The information collected about them included demographic parameters, general awareness of the pandemic, information about well-being, and anxiety score according to the GAD-7 scale (Saddik et al., 2021). The data were then processed using descriptive analysis to summarize characteristics, chi-square analysis to identify relationships between variables and levels of anxiety, and multivariable binary logistic regression analysis to identify preconditions.
The study made it possible to establish a relatively high level of anxiety in society at the survey time, especially among young people and the female population. Depending on which social group people belong to, the GAD-7 level of anxiety can be higher or lower. One way or another, the study’s main finding is the presence of a severe psychological impact of the pandemic on the population and a close relationship between the psychological state of children and their parents. Accordingly, there is an urgent need to introduce some screening and care measures for adults and children.
Given the nature of the research being conducted, chi-square statistic is one of the most appropriate methods for processing the available information. Firstly, this test is designed to compare the actual data obtained and the pre-existing model, which was carried out within the framework of this work. Second, the data used in chi-square should be random, which corresponds to the available random sample of volunteers. Finally, this method is specifically designed to test the dependence or independence of variables from each other, which was required to assess the psychological state and situation with COVID-19.
The method used is quite effective, and, besides, there are not many alternative approaches in such a situation. One of them is the Pearson correlation coefficient, which reveals the mathematical correlation between two variables. However, in such a situation, Pearson’s criterion has certain drawbacks: it is strictly mathematical, which requires additional data processing. In addition, this method shows only correlation, not showing causation. However, with minor modifications to the sampling of variables and their treatment, the Pearson correlation coefficient could be used in this study.
References
Li, H., Cui, L., Cao, L., Zhang, Y., Liu, Y., Deng, W., & Zhou, W. (2020). Identification of bipolar disorder using a combination of multimodality magnetic resonance imaging and machine learning techniques. BMC Psychiatry, 20(1), 1-12.
Saddik, B., Hussein, A., Albanna, A., Elbarazi, I., Al-Shujairi, A., Temsah, M.H., Sharif-Askari, F.S., Stip, E., Hamid, Q., & Halwani, R. (2021). The psychological impact of the COVID-19 pandemic on adults and children in the United Arab Emirates: a nationwide cross-sectional study. BMC Psychiatry, 21(1), 1-18.