PLAGIARISM FREE & "A" PLUS WORK IN 15 HOURS or LESS

Data Analysis Plan

Columbia Southern University

PUH 6301 Public Health Research

February 25, 2020

Data Analysis Plan

Checking for Data Accuracy

Data accuracy checking will incorporate various measures for efficacy. The first method will include using reliable data sources. The data sources are critical to successful data collection as well as further analysis. Therefore, I will ensure the credibility and reliability of the systems as well as personnel responsible for information and data generation. Another significant measure will be aligning the key parameters and factors. It entails analyzing and sifting through the features that contribute to data communication, by figuring out the most relevant parameters that are needed for the performance report of the specific operations or developing the feasibility (Cole & Trinh, 2017). Then, I will design a set of essential and basic parameters and formulate a plan for the data collection.

Equally, maintaining neutrality is essential for checking data accuracy since claims and exaggerations create a negative balance to data sets. Therefore, by ascertaining that data is neutral, it becomes easy to justify the completeness of data. Importantly, I will use computerized and automated programs. There is always room for more mistakes as well as a human error with the use of manual mechanisms during information recording and data entry (Cole & Trinh, 2017). Besides, there can be higher risks of inaccuracy and compromised data entries based on personal favors and biases that wholly affect data results and inferences, leading to loss of portability and efficacy of data accuracy and analysis. However, the data collection through automated and smart systems makes it easier for focusing on parameters and factors, while the system records accurate data and real-time in a perfect manner.

Level of Measurement

The important level of measurement for my research project is the nominal level of measurement. The measurement is essential to the research since it uses elements such as letters, words, numbers, and alpha-numeric (Ekinci, 2017). In the research, the hypothesis is establishing the difference in performance between private and public schools. Specifically, the hypothesis is “private schools perform better than public schools.” Therefore, one of the elements will be a comparison of performance by gender. In this case, female students will be classified as F and male students will be classified as M. The nominal level of measurement is equally essential in this research since it only possess the description of the character meaning the unique label for identifying values to subjects. In this case, it is used to identify male and female students and utilizes a one-on-one correlation between the objects and letters assigned. Therefore, the letters are merely for identifying the gender of the students and not their capabilities in the learning process.

Besides, the nominal level of measurement is essential in the research since it does not involve quantitative variables. Specifically, the research looks at the degree of performance across each school based on three selected items. 1= Higher Performance, 2= Moderate Performance, 3= Poor Performance. The number still, is only a representation of tags and not a value. Therefore, the nominal level measurement will apply to the different research elements applied in the research while comparing performance results between public and private schools (Ekinci, 2017).

Dependent Variables

I will collect dependent variables since performance can be influenced by a lot of factors. The dependent variable, in this case, is performance and is influenced by an independent variable such as curriculum. Learners have a similar curriculum, but performance will change based on the dependent factors (Ekinci, 2017). In this way, one of the variables I will collect concerning performance is how much study time is allocated for students in private schools versus those in public schools. Getting a difference in this variable will highlight the factors that contribute to the difference in performances. Also, I will look at the qualification of tutors among the schools. Specifically, it will involve looking at the institution with the most qualified personnel. In this way, it becomes easier to establish the impact of tutors on student’s performance in these different learning settings.

Besides, another important dependent variable I will measure is the availability of learning resources in public and private schools. It entails the identification of the learning institution with most resources as well as the least resources. Usually, schools with more resources will have better performances. Besides, another dependent variable to measure will include the social backgrounds of learners in each institution. Poor social background is associated with deprived learning incentives that generally affect student’s performance. Overall, by focusing on performance as the dependent variable, there are correlated variables that will result for investigation in the research.

T-test

The statistical test I would use it a T-test. It is important since it will be used as an inferential statistic to determine the important distinction between the means of two classes (the schools), which may be related to given features. The t-test is used to test the hypotheses and allows the testing of the assumption appropriate to the population (Schabenberger & Gotway, 2017). The t-test will take a sample from every two sets as well as ascertaining the problem statement by assuming that the null hypotheses are equal. Therefore, based on the relevant formulas, specific values are compared and calculated against the average values, and the hypothesis is rejected or accepted accordingly (Schabenberger & Gotway, 2017).

Therefore, this statistical analysis is essential since the outcomes of the test produce at-value. The t-value is compared alongside the value acquired from the critical value table. The comparison helps in the determination of the outcomes of the changes only on the variation, and whether the dissimilarity is outside the chance range (Schabenberger & Gotway, 2017). Hence, the t-test questions whether variations between the groups correspond to the true difference in the research or a worthless random difference. Besides, the t-test works when there is one independent variable with two categories, and there accompanying one dependent variable. Besides, the independent variable for the t-test is nominally scaled as established on the level of measurement.

Consulting an Expert

I will consult an expert since data analysis, specifically on the t-test. The data analysis and comparison of results should have higher accuracy, implying the involvement of an expert would be necessary. Besides, the expert can be essential in choosing the right data accuracy measure. Overall, an expert would be useful in the data analysis process by identifying the usefulness of the variable measurement and the level of measurements based on the research hypothesis. In most cases, an expert can offer an alternative research approach that fits the hypothesis based on the established variables and measurements mentioned (Schabenberger & Gotway, 2017). Importantly, the expert is a guide towards remaining on the correct track in data analysis, since he or she can offer the correct insights when there are errors or omissions in a section of the analysis.

References:

Cole, A. P., & Trinh, Q. D. (2017). Secondary data analysis: techniques for comparing interventions and their limitations. Current opinion in urology, 27(4), 354-359.

Ekinci, Y. (2017). Measurement of variables. Research methods for leisure, recreation and tourism, (Ed. 2), 77-95.

Schabenberger, O., & Gotway, C. A. (2017). Statistical methods for spatial data analysis. Chapman and Hall/CRC.