Analysis of Qualitative Data Set

Introduction 

As per the study by Lavie, Arena, and Blair, (2016), increasing weight is the rising issue of the 21st Century and various aspects have been observed as positively contributing to the increment of the weight among the new generation. Another study has contributed as, increasing weight is not just the issue of the new generation but some elder people have crossed the age of 50 and have spent approximately half of their lives, are also facing similar issues regarding their weight (Bowman, et al. 2016). Therefore, in such a case, it becomes necessary to analyze that what are the elements which are common in both the elders and the young generation. It was observed that most youngsters have been consuming junk food that contains a large amount of fat and it could be the main reason behind their increasing weight (Altaf, et al. 2018). On the other hand, another trend was analyzed that the consumption of junk food among the elders is quite low, which raises the question that what the other factors are significant in this manner.

Furthermore, technological advancement has taken place which simplifies the job of the daily routine and it is the main reason behind the physical inactivity of most people. Moreover, it was studied that the lifestyle that a person has been availing is dependent on their income, which makes it clear that if a person is on a higher level of income, there are chances that he would be using more technological aspects in his daily life. This will directly contribute to his physical inactivity and it is evident by the study by Myers, et al. (2017), that if a person is physically inactive, there are chances that the weight of the person would increase significantly irrespective of his age and other factors. However, to make it clear that the aspects which contribute to the increasing weight are significant, this study has been conducted in which various independent variables have been analyzed to observe their impact on the single dependent variable which is the weight of the people.

Jakoi, et al. (2017), have mentioned in their study that there is a significant relationship between smoking status and the increasing weight of the body. However, various examples have been observed in which people who smoke were observed as losing weight. Therefore, in such a case, it was necessary to include the smoking status of the people in this study so that it would be clear if the smoking habit of people increases their weight or not. Factors that have been taken as the independent variable in this study include age, income level, height and weight of the body, gender, and smoking status. To keep the findings of the study unbiased, the data has been taken from a large number of sampthathich appropriathatrepresents the represent. Based upon the selected sample and methodology tests were applied to obtain the findings which helped in presenting the comprehensive conclusion

Methods and Materials

To conduct any study, the most important element is to select the right method, as it has been studied that most studies fail due to the weak methodology. There is a various aspect which is necessary to analyze while selecting the methodology for any study such as the selection of the sample, selecting the right tests, and the factors which should affect the dependent variable of the study. While selecting the sample size, it is necessary to analyze how large the population of the study is. Therefore, analyzing the population of the study, a sample size should be selected, it was observed study by Etikan, and Bala, (2017), that if the population is too large, it is necessary to keep the sample size also large so that the findings of the study would be unbiased. Furthermore, once the sample size of the study has been selected, the research instrument should be decided what would be the method for collecting the data from the sample (Taherdoost, 2016). Therefore, in this study, the data has been collected with the help of a survey questionnaire, in which the respondents of the study were provided with closed-ended questions and they had to select the given options regarding the different factors such as their gender and smoking status. However, for weight, height, and age, they were allowed to enter their respective answers in the given unit such as pound, inches, and years respectively. Moreover, when the respondents were asked to provide details regarding their income, they were provided with different slabs which define the income level and they had select any of the brackets in which their income lies.

Based upon the selected data, and the aim of the study, in which the researcher intended to find the impact of the multiple independent variables over a single dependent variable, test of the study was defined. Therefore, analyzing the related elements, multiple regression was applied over the sample of the study. Moreover, it is necessary to mention that the sample was taken from a large number of respondents that is 1350 people.  Furthermore, while collecting the data for the study, there were some limitations tthatwasecided as the respondents should be above the age of 20, therefore, the respondents are between the ages of 20 to 89 years. In addition, their income brackets were defined into various levels between less than USD 1,000 to more than USD 150,000. The reason to break the income level into various brackets was to categorize the living standards based on their income levels as it has been studied that the income should be divided into levels to simplify the prediction of the tests’ results (Padley, and Marshall, 2018). 

In addition, the most important aspect of the methodology of the study is, what approach has been used in the study. Two types of approaches can be used in any study, such as inductive and deductive approaches, and the decision for the approach of the study is usually made upon the fact that the study is investigating the results of the previously developed theories or developing its own (Jebb, Parrigon, and Woo, 2017). In this study, the researcher intended to develop its theory including the impact of the income level, age, height, gender, and smoking status of the people over their weight. Therefore, in this study, the researcher observed the patterns from the related literature that what are the factors which contribute to the increasing weight of the body and based upon it, tentative hypotheses were made. Furthermore, once the data was collected and the tests were applied, results were provided to the researcher which helped in making the conclusion which is considered the theory that has been developed by the study with the help of primary data collected by the researcher. In addition, to maintain the level of ethics in the study, every piece of information which is taken from the other studies has been credited to the real author of the study (Pietilä, et al. 2020). Furthermore, while taking the information from the secondary resources, the authenticity of the source has been considered so that the findings of the study should not be biased.

Results

This section of the assignment includes the results that have been obtained by executing relevant tests on SPSS on the given data set to reach suitable findings and conclusions. This section is divided into different parts to explain the findings properly.

Testing for Assumptions

Normality

The data set provided is assumed to follow normal distribution. For the purpose of assessing the normality of each variable that has been taken into consideration, Kolmogorov-Smirnov test and Shapiro-Wilk test have been applied. The null hypothesis for this test is that the data is normally distributed, in case if the sig value are less than 0.05 or alpha value it indicates that the data is not normally distributed. The following table shows the normality for each variable taken into consideration:

Table 1: Normality Test

Source: Author (2020)

In the above table, it becomes evident that for all the variables the values for the Kolmogorov-Smirnov test and Shapiro-Wilk test appear to be 0.000 which is less than the benchmark set which indicates that the null hypothesis is deemed to be rejected. Hence, the dataset is not normally distributed. 

Multicollinearity

Another major assumption for the data set is that there is no existence of multicollinearity in the existing dataset. In the case of the existence of multicollinearity, the statistical significance of the data is diluted hence, it is assumed that the data does not have the problem of multicollinearity. The following table shows the value of the Variance Inflation Factor (VIF) for each variable which needs to be less than 10 to prove that there is no multicollinearity in the data:

Table 2: Multicollinearity Test

Source: Author (2020)

It is evident from the above table that there is no existence of the problem of multicollinearity in the variables since all the VIF values are less than 10. 

Linear Relationship

Before testing different attributes for the weight of the sample size, it has been assumed that all the independent variables have a linear relationship with the dependent variable which is weight in pounds. To test this, each of the predicting variables is plotted against the weight to visually analyze the linearity, following are the graphs:

The above table indicates that there is a linear relationship between the weight of an individual and their height. The relation is seen to be positive as the line is upwards which means an increase in height brings about an increase in weight as well. 

The above table indicates that there is no linear relationship between total family income and weight. It can be seen in the above graph that there are many fluctuations in the data which means that there is no linear pattern. 

The above table indicates that there is no linear relationship between the age of an individual and their weight. It can be seen in the above graph that there are many fluctuations in the data which means that there is no linear pattern. 

Multiple Regression

After testing the data for assumptions, multiple regression has been applied to the data set to find out the influence of predictors (age, gender, smoking status, height, and family income) on the criterion variable (weight in pounds). As mentioned in the methods and materials section, this data has been analyzed via SPSS. Considering that there were several independent variables in the model, the researcher applied multiple regression to test the relationship and impact. The following table indicates the results that have been obtained from SPSS: 

Table 3: Ordinary Least Squares

Source: Author (2020)

The above table shows all the relevant information that is required to analyze the significance and extent of the association/relationship of different attributes with the weight of individuals. The top portion of the table indicates the coefficient values and their respective sig values which shows whether or not there is statistical relevance in their relationship. Coefficient values for each of the predicting variables indicate the difference that is apparent in the dependent value due to one unit change in them. 

The first predicting variable is Age which is tested against the weight in pounds of the respondents. The p-value assigned to this variable is 0.551 which lies outside the benchmark of 0.05 or below, thus, the individualistic impact of age on the weight of individuals is found to be statistically irrelevant. Considering that the relationship is insignificant, there is no point in explaining the coefficient value since it depicts the extent of change in the weight caused by one unit change in age. 

Secondly, the impact of gender is tested on the weight of an individual, which means that whether or not belonging to a specific gender influences the weight of a person. . The p-value assigned to this variable is 0.003 which lies within the benchmark of 0.05 or below, thus, the individualistic impact of gender on the weight of individuals is found to be statistically relevant and significant. The coefficient value for this variable is -8.590 which means that gender can significantly predict the weight of an individual. 

The next variable is smoking status which means whether or not an individual smokes. The impact of this variable is tested on the weight of an individual, this means whether or not smoking influences the weight of a person. The p-value assigned to this variable is 0.000 which lies within the benchmark of 0.05 or below, thus, the individualistic impact of an individual being a smoker on the weight of individuals is found to be statistically relevant and significant. The coefficient value for this variable is 11.195 which means that smoking status can significantly predict the weight of an individual.

The next individual attribute that has been tested is height in inches. The impact of this variable is tested on the weight of an individual, this means that whether a person is taller or shorter impacts their weight or not. The p-value assigned to this variable is 0.000 which lies within the benchmark of 0.05 or below, thus, the individualistic impact of the height of an individual on the weight of individuals is found to be statistically relevant and significant. The coefficient value for this variable is 4.663, this indicates that if an inch of height of an individual is increased, it will bring about a 4.663 units change in the weight of a person. 

The last variable is total family income which is not a personal attribute but a social factor. The impact of this variable is tested on the weight of an individual, this means whether a person belonging to a certain income group impacts their weight or not. The p-value assigned to this variable is 0.011 which lies within the benchmark of 0.05 or below, thus, the individualistic impact of total family income on the weight of individuals is found to be statistically relevant and significant. The coefficient value for this variable is -0.504, this indicates that an increase in total family income can reduce the weight of an individual as there is a negative sign associated with the coefficient. 

In the bottom section of the table, certain values collectively explain the entire model including all the predicting factors and crvariablesvariable taken into consideration. Sig. value determined the complete statistical relevance of the model. The p-value assigned to the model is 0.000 which lies within the benchmark of 0.05 or below, thus, the overall and collective impact of total family income, age, gender, height, and smoking on the weight of individuals is found to be statistically relevant and significant. R-square value is another important value to be interpreted to explain ordinary least squares regression in a detailed manner. R-square value otherwise known as the coefficient of determination shows the ability of the independent variable in terms of predicting the variance that occurs in the dependent variable. In the case of the model in consideration, total family income, age, gender, height, and smoking can explain 0.252 or 25.2% of the variation in the weight of respondents. Lastly, an adjusted R-square can be referred to as a modified version of the coefficient of determination which is being adjusted by the number of independent variables in the model. Thus, after adjustment, total family income, age, gender, height, and smoking can explain 0.249 or 24.9% of the variation in weight of individuals.