Critically compare the statistical material in this document with items 10-12 of the STROBE list.
Use the data set you were given for your Assignments to answer the following research question.
Is the logarithm for MVPA able to predict GPA after correcting the overweight population of Australian university students
Your evaluation of the strengths and limitations of the presentation of statistical material in Fox et. al. 2010 against items 10-12-17 of STROBE
Researchers used 4746 high school and middle school students as a sample.
It is unclear how the sampling was done.
Researchers have not provided details about the sampling methods used, including whether they used simple random, cluster, systematic, or stratified sampling.
It is crucial to explain the methods used to collect data so that respondents can make generalizations.
I think the researchers did a great job in providing demographic information about the respondents.
They gave us information such as the gender ratio, age brackets, ethnicity, and socio-economic status of students.
Researchers have not provided any information regarding control for possible confounding that may have occurred during the study.
It is assumed that no confounders were necessary to prevent biased results.
Researchers did not mention, but rather pointed out, whether there were any issues with missing data.
Missing data, also known as missing values, is when no data value has been stored for one or more variables within an observation.
Missing data can be a significant problem in research and can impact the conclusions and results.
Parametric tests were conducted for inferential analysis when it came down to analysis.
Their test of the assumptions required for parametric tests is missing.
Parametric tests cannot be distributed free. They must conform to certain assumptions. If any assumptions are violated, biased results may result.
Parametric tests are subject to certain assumptions, including normality assumption, linearity assumption and homogeneity assumption (also called equal variances assumption or independence assumption).
Present your descriptive analyses
This question sought to answer the question: Does the logarithm for MVPA predict GPA? After correcting for overweight among Australian university students.
The summary statistics for both variables were first examined.
The logMVPA average was 0.44, while the median was 0.52 (a little higher than the mean).
The lowest value was -0.52, while the maximum value was 1.22, with a range between 1.75 and 1.22.
Both the skewness (-0.49) and the kurtosis (-0.46) are negative values. This could indicate that data is negatively skewed.
Figure 1 is below.
Figure 1: Histogram of logMVPA
Students had an average GPA score score of 4.76 and a median score score of 4.7.
This group had a 6.9 GPA and a 2.4 GPA.
Summary statistics also showed the values for skewness (-0.08) and kurtosis (-0.44) respectively.
The skewness value being close to zero indicates that the data came from a normally distributed set.
The histogram in figure 2 below confirms this notion.
The histogram of the GPA is shown in Figure 2.
The figure clearly shows that data for variable GPA are normally distributed (bell-shaped curve).
Figure 2: GPA Histogram
Present the results of relevant regression models or inferential analyses (about 150-200 Words, 10 Marks).
We ran a linear regression model and a Pearson correlation test to see if logMVPA can predict GPA.
The Pearson correlation test revealed that the coefficient between logMVPA and GPA was 0.6648. This indicates that there is a strong positive relationship between logMVPA and GPA.
The scatter plot of GPA against logMVPA, which accounts for overweight, is also shown.
Figure 3: A scatter plot of GPA/logMVPA
This figure clearly shows the positive linear relationship between GPA (Grade Point Average) and logMVPA.
A regression analysis was performed to arrive at a model that could predict GPA using the logarithms of the MVPA (logMVPA).
We had to examine two aspects of the model’s fit in order to determine its fitness. These were the coefficient of determination (or significance value) and the model’s magnitude.
The goodness of the fit of the model was first examined. It is able to predict GPA using logMVPA at 5% significance (p 0.05).
R-squared is the coefficient of determination. This means that 44.2% variation in dependent variables (GPA) can be explained by the model’s explanatory variable logMVPA.
LogMVPA, a variable that was significant in the model (p > 0.050), had a coefficient of 1.6256. This means that a one-unit change in logMVPA would cause a change to the GPA by 1.66256.
If logMVPA rises by 1 unit, then we expect that the GPA will increase by 1.6256.
Similar to the logMVPA, GPA would also decrease by 1 unit if it falls by 1.62256.
The constant intercept was 4.3387
Answer the research question.
This study was done to determine if the logarithm (logMVPA), of MVPA (logMVPA), predicts GPA after correcting overweight among Australian university students.
To answer this question, regression models were constructed.
The logMVPA can predict the GPA when overweight is controlled.