After I got the data, I created the following two tables in order to draw the scatter plots:
It’s not necessary to include the students’ name because their names are not related to my experiment, but their gender is required because male and female may have different relationship between their humerus length and height and I need to 2 scatter plots in order to have more detailed results.
I used my data to draw 3 scatters including the line of best fit ( ax + b ). Here are the 3 scatter plots:
Analysis of Data
After I created the graphs, I found that according the line of best fit in the 3 graphs. People’s height is increased along with their humerus length for both male and female. And finally I created a graph, which is combined the data of male and female into one graph. I still found out the same overall pattern; people’s height is creased along with their humerus length. So, I made a wrong assumption in the experiment. The relationship between humerus length and height of people won’t be really affected by the gender of people.
Errors or Bias
I used the convenience-sampling method to get the sample. A group of students can’t reflect the overall people in the world. There is a bias that there aren’t enough observations in my sample.
The people who measured all people’s height in my class only use a ruler to put on each person’s head to get their height. The person who measured people’s height may not reach the horizontal line when he put the ruler on the others’ heads. Meanwhile, people wear shoes when their heights were being measured, and that would affect the result.
Finally, students measured their humerus length by themselves; they didn’t really understand the concept of humerus length. People measured from different positions, for example, some people measured from the end of the shoulders to the beginning of the elbows, but some measured from the beginning of the shoulders to the end of the elbows.
Ways to improve the experiment
I suggested the following ways to improve the experiment:
- Only one person measure everyone’s humerus length
- Use mechanical machine to measure people’s height and only one person saying out the results to reduce bias
- Do not wear shoes when heights are being measured
- Having more observations, for example, having 2-3 classes to do the experiment at the same time
Conclusion
Overall, I think this experiment was a failure because there was too many errors and bias in this experiment. The observations of this survey are absolutely not enough. Also, the people didn’t measure their humerus length correctly, which would affect the result of this survey very much.