Research Example: Statistical Analysis Whether People's Weight Depends on Their Lifestyle/Diet

Statement of Intent

The objective of this research is to determine by use of statistical analysis whether people's weight depends on their lifestyle/diet. The finding from the research process will enable me to actually know whether there is a correlation between diet and weight. This will prove important in helping me come up with a concrete statement on the two during the Universities Nutrition Day. The data will also help me in coming up with a plan on how to advise people on their lifestyle/diet and how this impacts their weight issues. I will use the findings to back up my advice to them as they will be statistically valid following the statistical significance tests that I will have used during the research.

For the purpose of this research, the data will be based on a survey conducted on 200 University Students weighing between 55 and 80 kilograms. The sample population will be chosen randomly from their halls of residence. This will be carried out during end of the academic period to ensure that almost everyone is within the school's premise.

As part of discovering the correlation between people's weight and their diet, I will employ statistical tool such as scatter plots and use of statistical significance test based on selected hypothesis and regression analysis. This will ensure that the findings from the study is statistically significance and valid for use based on the mathematical accuracy.

Raw DataJust like any other giver research process, the data collection is a crucial process (Corder & Foreman, 2014). It determines the accuracy of the statistical test. In light of this truth, it was necessary to come up with a method of collecting data that would ensure that the process was carried out in a manner that eliminates all form of bias. The collection of the data was therefore, carried out in a random manner as part of eliminating any form of confirmatory bias which serves a big source of errors. For this research on whether there exists a correlation between plant-based diet/meat diet and people's weight, data was obtained from a series of questions and prompts for the purpose of gathering information on people's weights and diets they take. This was conducted in a friendly manner that was easily understood by the people as part of minimizing any form of bias. It was also conducted in an ethical manner by explaining to the people the purpose of the study. They were informed that this information was to be used only for the purpose stated and not in any other way.

Weight plant based meat based total

55 6 1 7

56 3 2 5

57 3 0 3

58 4 4 8

59 5 6 11

60 5 4 9

61 6 0 6

62 3 1 4

63 3 5 8

64 0 2 2

65 1 2 3

66 3 6 9

67 3 2 5

68 6 3 9

69 10 4 14

70 5 5 10

71 15 11 26

72 8 4 12

73 5 5 10

74 3 4 7

75 3 6 9

76 1 0 1

77 5 1 6

78 0 4 4

79 1 3 4

80 2 6 8

200

Table 1 showing raw data for the experiment

The data collected was divided into three parts. There was information on the weight of the individuals, the information on those taking plant-based diet and those taking meat-based diet. The information was statistically enough for sampling purpose as it serves to reflect information about the given population under study (Rosner, 2015). The above shown is the raw data for the research process obtained from the survey process used to collect the data.

Scatter plots

The use of scatter plots in a given research process makes it easier for one to analyze results from a given statistical test at a glance (Le & Eberly, 2016). They help to show the trends in the findings obtained from the study. It is in light of this recognition that my study made use of scatter plots. Their use was to show the trends in weights and the diet taken by the students. By taking a closer look at the plots, it is easier to identify whether there is a correlation between the variables used in the study. The x-axis of the scatter plot graph was the weight of the students in kilograms while the y-axis was the diet taken by the students.

Based on the observation from the scatter plot, it is evident that there is no clear trend between the weights of the sample population and diets. It is out of this observation that prompts me to further carry out studies on the variables to see whether a correlation coefficient will prove whether there is a strong linear relationship between the two variables under study.

Correlation Coefficient

The correlation coefficient is a statistical coefficient used to determine whether there is a strong linear relationship between two variables under study (Vittinghoff et al., 2014). The following is the determination of whether there exist a strong linear correlation between the weight of the students and the plant-based diet.

For the weight and plant-based diet

Based on the calculations, the following was obtained:

= 67.5

= 4.2

= 119925

=717

= 7302

n=26

The correlation coefficient r obtained in the calculation was -0.09. This shows a weak correlation between the weight of the students and the plant-based diet. In coming up with this conclusion, the fact that a closer value of r is to +1, a strong correlation is inferred (Vittinghoff et al., 2014). If the r is closer to -1, there is a stronger negative correlation. For the case of the study, it was obtained that the r-value was closer to Zero hence the poor/weak correlation between the two variables. It is also important to note that there is a negative correlation showing that an increase in the plant-based diet leads to a decrease in weight of the students.

For weight and meat-based diet

Having obtained the correlation coefficient for the relationship between weight of the participants and the plant-based diet, it was prudent to check the relationship between the weight and the meat-based diet. The same formula for calculating the correlation coefficient r was used as had been used in the case of the previous calculation. The following was obtained:

=473

= 6274

Based on the calculations, the value of the correlation coefficient r was found to be 0.02766. It was a positive value in comparison to the one obtained in the determination of the correlation between the weight and the plant-based diet. The value (r=0.2766) was indicative of a weak correlation between weight and diet. It is however prudent to note that for this case, the correlation is positive thus indicating that any increase in uptake of meat-based diet will significantly lead to an increase in the weight of the individual.

Testing the hypothesis using an ANOVA TestsAs part of further confirming the test to determine their statistical significance, I subjected the variables to an ANOVA test to further confirm the relationship between the weight and the diet. The ANOVA test was guided by the following hypothesis:

For weight and plant-based diet

Hypothesis

H0: No relationship between weight and diet

H1: There exists a relationship between weight and diet

Anova output from excel

ANOVA Source of Variation SS dfMS F P-value F crit

Between Groups 52102.23 1 52102.23 1512.368 4.86E-39 4.03431

Within Groups 1722.538 50 34.45077 Total 53824.77 51

From the output F (Observed) = 1512.368> F (Critical) = 4.03, we therefore reject H0 and accept H1: This leads to a conclusion that there exist a significant relationship between the two variables in question

For weight and meat-based diet

Hypothesis

H0: No relationship between weight and diet

H1: There exist a relationship between weight and diet

Anova output from excel

ANOVA Source of Variation SS dfMS F P-value F critBetween Groups 53248 1 53248 1646.506 6.19E-40 4.03431

Within Groups 1617 50 32.34 Total 54865 51

From the output F (Observed) = 1646.506> F (Critical) = 4.03,

We therefore, reject H0 and accept H1. This leads to a conclusion that there exist significant relationships between the two variables in question

CONCLUSION

Based on the research finding, it was clearly noted that there exist a weak correlation between weights and lifestyle/diet as noted in the values of r obtained from the calculations. It is however noteworthy to realize that a strong relationship between people's weight and their lifestyles from the ANOVA tests conducted on the University students.

Furthermore, using the standard deviations point of view, one can explore the relationship between the variables in question as part of testing their stability. The standard deviation for the plant-based diet data is 3.23 while that of meat-based data is 2.49. This shows that there is a high stability in weight for those who take meat-based diet. This is because its standard deviation is small compared to those of the plant-based diet. The finding can thus help in coming up with a comprehensive report for use in the University Nutrition Week as my data is statistically valid.

References

Corder, G. W., & Foreman, D. I. (2014). Nonparametric statistics: A step-by-step approach.

John Wiley & Sons.Le, C. T., & Eberly, L. E. (2016). Introductory biostatistics.

John Wiley & Sons.Rosner, B. (2015). Fundamentals of biostatistics.

Nelson Education.Scott, I., & Mazhindu, D. (2014). Statistics for healthcare professionals: An introduction. Sage.

Vittinghoff, E., Glidden, D. V., Shiboski, S. C., & McCulloch, C. E. (2011). Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. Springer Science & Business Media.