The Project:
For this project, we had to figure out a way to determine someone's height based on their gait. A gait is a person's manner of walking. We initially conducted trials whereby we calculated average step lengths for individual team members of varying heights. From that, we were able to use those heights to determine a coefficient. With that coefficient, we then developed a predictive formula: average step distance (in feet) divided by the coefficient equals a person's estimated height (in feet). After we found this formula, we tested it to see if it was indeed accurately predictive. We realized, after plotting our data, that we had more outliers than anticipated. Accordingly, we changed our graph to find a better line of best fit. We found it would be better to use a root ten graph than a linear graph. Our final predictive formula was y = 10√(-8.8*107 + 4.9*107x), which had a strong 87% accuracy.
Lab Write-Up:
Micro Presentation:
Content:
Gait: The stride of a human as they move their limbs.
Predictive modeling: A process that uses data mining and probability to forecast outcomes. In this project, our predictive model was an equation.
Accelerometer: An instrument for measuring acceleration. The accelerometer measured our accelerations in three different dimensions:
Variability: The quantification of fluctuations from one stride to the next. Variability can be seen in our graphs by comparing the sizes of each wave.
Predictive modeling: A process that uses data mining and probability to forecast outcomes. In this project, our predictive model was an equation.
Accelerometer: An instrument for measuring acceleration. The accelerometer measured our accelerations in three different dimensions:
- Vertical - up and down direction
- Lateral - side to side direction
- Anterior - front and back direction
Variability: The quantification of fluctuations from one stride to the next. Variability can be seen in our graphs by comparing the sizes of each wave.
Reflection:
This project was one of the more confusing projects we've encountered to date. We had to use an accelerometer app, and it was difficult to interpret because it spewed out so much data all at once and in three different dimensions; we were not entirely sure what the dimensions were referring to. On the upside, the large amounts of data required us to become skilled at putting data into spreadsheets, and from those spreadsheets, we were able to create graphs and predictive models.
We approached this project a little differently from peer groups. Most worked directly from the graphs given from the accelerometer app (and took the equation directly from the graph given); our group chose to conduct various trials with an equation that we produced ourselves using the information we gained through our experiments. Although we didn’t end up using our original equation in the long run, it was good that we took the path less travelled, as we felt that it gave us a deeper understanding of the reasoning behind the equation. This made the project more satisfying than usual.
We approached this project a little differently from peer groups. Most worked directly from the graphs given from the accelerometer app (and took the equation directly from the graph given); our group chose to conduct various trials with an equation that we produced ourselves using the information we gained through our experiments. Although we didn’t end up using our original equation in the long run, it was good that we took the path less travelled, as we felt that it gave us a deeper understanding of the reasoning behind the equation. This made the project more satisfying than usual.