We have chosen the topic "Is a Person's BMI Related to his Height?". This study is important because height is used in the calculation of the BMI.

The variables for this research topic are:

1. Independent variable: Height

2. Dependent variable: BMI

In this research, height means how tall a person is. The unit of measurement is metre (m).

BMI means Body Mass Index. It is a tool to measure whether a person is underweight, normal, overweight or obese, based on his body weight and height. BMI is measured using the following formula:

BMI = Weight / Height x Height

(note: unit of weight is kg)

Before we embarked on collecting data, we determined the hypothesis for this study.

H0 : A person's BMI is not related to his height.

H1 : A person's BMI is related to his height.

At the end of our study, we shall see if the hypothesis is to be rejected or fail to be rejected.

## Thursday, August 13, 2009

## Wednesday, August 12, 2009

### 2. Methodology

On Friday 10th july 2009, we gathered at NYP SHS Block J to collect data for our study. The data collection procedure was as follows:

1. A digital weighing scale was set up.

2. A tape measure was set up against a door, and was made to be as perpendicular to the floor as possible.

3. By using only one weighing balance and one tape measure for all subjects, standardization was ensured.

4. Then, we gathered 30 willing subjects who gave their verbal consent to have their weights and heights taken.

5. Each subject had their weight and height taken 3 times.

6. The readings were written down and then entered into SPSS for further processing.

To measure the weight, each subject stood on the digital weighing scale 3 times and an appointed group member jotted down each measurement. Another group member was also there to double check that each reading was read correctly.

On the other hand, to measure the height, each subject was instructed to stand with his back straight against the tape measure that was set up, and an appointed group member measured the height from base of feet to top of head. Once again, an extra group member was there to double check that the reading was read correctly. This procedure was repeated 3 times.

All these readings were entered in to SPSS and an average of each subject's 3 weight readings and 3 height readings were calculated and ready to be analysed.

Note: No person rejected when approached to be a subject for this study.

Note: No reading was rejected.

These are the photos taken during the data collection. Enjoy.

1. A digital weighing scale was set up.

2. A tape measure was set up against a door, and was made to be as perpendicular to the floor as possible.

3. By using only one weighing balance and one tape measure for all subjects, standardization was ensured.

4. Then, we gathered 30 willing subjects who gave their verbal consent to have their weights and heights taken.

5. Each subject had their weight and height taken 3 times.

6. The readings were written down and then entered into SPSS for further processing.

To measure the weight, each subject stood on the digital weighing scale 3 times and an appointed group member jotted down each measurement. Another group member was also there to double check that each reading was read correctly.

On the other hand, to measure the height, each subject was instructed to stand with his back straight against the tape measure that was set up, and an appointed group member measured the height from base of feet to top of head. Once again, an extra group member was there to double check that the reading was read correctly. This procedure was repeated 3 times.

All these readings were entered in to SPSS and an average of each subject's 3 weight readings and 3 height readings were calculated and ready to be analysed.

Note: No person rejected when approached to be a subject for this study.

Note: No reading was rejected.

These are the photos taken during the data collection. Enjoy.

## Tuesday, August 11, 2009

### 3. Data collected

The images above are the data collected on Friday, 10th July 2009. Click on the images for a bigger view.

The 1st image is the data collected.

The 2nd image is the details of each variable.

Sex: sex of respondent

wt: Weight (kg)

ht: Height (m)

avwt: Average weight (kg)

avht: Average height (m)

BMI: Body Mass Index (weight / height x height)

For those who are interested, below is the BMI table obtained from the World Health Organization (WHO). Click for a bigger view.

Note: For all images in subsequent blogposts, please click for a larger view.

## Monday, August 10, 2009

### 4. Appropriate Statistical Test.

Before we start analyzing the data, first we have to determine which statistical test we are going to use.

Choosing the appropriate statistical test depends on the type of research question, whether is it

1. Difference, or

2. Correlation.

For our study, the research question is a correlation question, i.e. "Is a person's BMI related to his height?".

Another factor that we have to consider is the type of measurement of our variables.

Both the independent (height) and dependent (BMI) variables are scale variables

Taking reference from Figure 8.2 "Decision path - Relationship Question", Statistics in Health Sciences, 4th Edition, by Chia Choon Yee, we have determined that the appropriate statistical test to use for our study is Pearson's r.

Now that we have determined that Pearson's r is the statistical test we are going to use, here comes the big question. What is Pearson's r?

Pearson's r is a symmetric measure of association for interval level variables. Pearson's correlation coefficient ranges from -1.0 to +1.0

+1.0 indicates a perfect positive relationship, while,

-1.0 indicates a perfect negative relationship.

When using Pearson's r, there are four assumptions we must take note of.

Choosing the appropriate statistical test depends on the type of research question, whether is it

1. Difference, or

2. Correlation.

For our study, the research question is a correlation question, i.e. "Is a person's BMI related to his height?".

Another factor that we have to consider is the type of measurement of our variables.

Both the independent (height) and dependent (BMI) variables are scale variables

Taking reference from Figure 8.2 "Decision path - Relationship Question", Statistics in Health Sciences, 4th Edition, by Chia Choon Yee, we have determined that the appropriate statistical test to use for our study is Pearson's r.

Now that we have determined that Pearson's r is the statistical test we are going to use, here comes the big question. What is Pearson's r?

Pearson's r is a symmetric measure of association for interval level variables. Pearson's correlation coefficient ranges from -1.0 to +1.0

+1.0 indicates a perfect positive relationship, while,

-1.0 indicates a perfect negative relationship.

When using Pearson's r, there are four assumptions we must take note of.

- Assumption 1 : All observations must be independent of each other.
- Assumption 2 : The dependent variable should be normally distributed at each value of the independent variable.
- Assumption 3 : The dependent variable should have the same variablility at each value of the independent variable.
- Assumption 4 : The relationship between the dependent and independent variables should be linear.

## Sunday, August 9, 2009

### 5. Data Analysis (part A)

Research question: Is a person's BMI related to his height?

H0: A person's BMI is not related to his height

H1: A person's BMI is related to his height.

Firstly, we generated a scatter plot to check for linearity and homogenous variance. This is to test for assumptions 3 and 4 of Pearson's r.

The scatter plot appears to follow a general positive linear trend, though the relationship is very weak. However, there is no violation of the linearity assumption.

Following that, we compute Pearson's r.

From the table above, Pearson's r value is 0.163. This shows that there is a positive but very weak relationship between BMI and height.

The p value is 0.391, which is more than 0.05. When p > 0.05, we fail to reject the null hypothesis (H0).

In conclusion, a person's BMI is not related to his height.

H0: A person's BMI is not related to his height

H1: A person's BMI is related to his height.

Firstly, we generated a scatter plot to check for linearity and homogenous variance. This is to test for assumptions 3 and 4 of Pearson's r.

The scatter plot appears to follow a general positive linear trend, though the relationship is very weak. However, there is no violation of the linearity assumption.

Following that, we compute Pearson's r.

From the table above, Pearson's r value is 0.163. This shows that there is a positive but very weak relationship between BMI and height.

The p value is 0.391, which is more than 0.05. When p > 0.05, we fail to reject the null hypothesis (H0).

In conclusion, a person's BMI is not related to his height.

## Saturday, August 8, 2009

### 6. Data Analysis (Part B)

Since we found that a person's BMI is not related to his height, we have decided to explore a little further and analyze if a person's BMI is related to his weight.

Research question: Is a person's BMI related to his weight?

H0: A person's BMI is not related to his weight.

H1: A person's BMI is related to his weight.

A scatter plot was generated.

The scatter plot appears to follow a general positive linear trend, and there is no violation of the linearity assumption.

Then we compute the Pearson's r.

From the table above, Pearson's r value is 0.880. This shows that there is a positive and very strong relationship between BMI and weight. The p value is 0.000, which is less than 0.05.

When p is less than or equal to 0.05 we reject null hypothesis.

In conclusion, a person's BMI is related to his weight.

Research question: Is a person's BMI related to his weight?

H0: A person's BMI is not related to his weight.

H1: A person's BMI is related to his weight.

A scatter plot was generated.

The scatter plot appears to follow a general positive linear trend, and there is no violation of the linearity assumption.

Then we compute the Pearson's r.

From the table above, Pearson's r value is 0.880. This shows that there is a positive and very strong relationship between BMI and weight. The p value is 0.000, which is less than 0.05.

When p is less than or equal to 0.05 we reject null hypothesis.

In conclusion, a person's BMI is related to his weight.

## Friday, August 7, 2009

### 7. Reflections

At first glance, the project seemed relatively easy to carry out. But as we were to find out as we journeyed on, the path was speckled with obstacles here and there.

The journey can be divided into two halves. The collection of data and the analysis of data.

The first half, the collection of data, was fun and enjoyable. We had no problem deciding which was the dependent and independent variables, and was planning so enthusiastically on how to carry out the measurements needed for our little study. We learned that teamwork was particularly important to ensure a smooth process during data collection.

Then the second half of the journey commenced! No more fun and frolic, but time to sit down and rack our brains on how to analyze our data correctly. There were many things to consider, such as the level of measurement of our variables, which statistical test to use, how to interpret the results and so on. Things weren't as easy as it was perceived to be and we realised that statistics is NOT a magic show where we can click a few times here, click a few times there and voila!, everything is done! No, not at all! A lot of planning, understanding and application of knowledge was required, and we are so thankful for the wonderful lectures and tutorials that really aided us in this project!

Well, here we are, at the end of our journey. There are so many people we want to thank, and no one we would want to hate, of course! We all cherish every single up and down moment that we experienced and who would dare deny that we had much much more "up" moments than "down" moments.

All in all, this journey opened up our eyes to the wonderful world of statistics and we bring home nothing but sweet memories and better-filled brains. Oh, and not to mention, we all learnt how to blog!

Have a happy day!

love,

the Beautiful People.

The journey can be divided into two halves. The collection of data and the analysis of data.

The first half, the collection of data, was fun and enjoyable. We had no problem deciding which was the dependent and independent variables, and was planning so enthusiastically on how to carry out the measurements needed for our little study. We learned that teamwork was particularly important to ensure a smooth process during data collection.

Then the second half of the journey commenced! No more fun and frolic, but time to sit down and rack our brains on how to analyze our data correctly. There were many things to consider, such as the level of measurement of our variables, which statistical test to use, how to interpret the results and so on. Things weren't as easy as it was perceived to be and we realised that statistics is NOT a magic show where we can click a few times here, click a few times there and voila!, everything is done! No, not at all! A lot of planning, understanding and application of knowledge was required, and we are so thankful for the wonderful lectures and tutorials that really aided us in this project!

Well, here we are, at the end of our journey. There are so many people we want to thank, and no one we would want to hate, of course! We all cherish every single up and down moment that we experienced and who would dare deny that we had much much more "up" moments than "down" moments.

All in all, this journey opened up our eyes to the wonderful world of statistics and we bring home nothing but sweet memories and better-filled brains. Oh, and not to mention, we all learnt how to blog!

Have a happy day!

love,

the Beautiful People.

Subscribe to:
Posts (Atom)