Household Water Usage by Amalia Lazo-Navarrete

Amalia Lazo

amalialazonavarrete — 2024-03-16 05:04:47 UTC

The purpose of this project was to compare water usage to the number of people in a household in other states, our households, and our city. The background information was collected by using data from Google sheets, cleaning it up, and then putting it in stapplet and codap in order to get the R2 and LSRL. For my claim I said that where you live like neighborhood and city would impact your water usage because there might be water shortages in some locations or some rules enforced in order to conserve water.

Scatterplot

amalialazonavarrete — 2024-03-16 05:14:20 UTC

Questions

amalialazonavarrete — 2024-03-16 05:14:51 UTC

What I see from the scatterplot is how there are two types of variables and how their relationship is being measured by the line in the middle to see how well they compare to each other.
The y-intercept in the context of the line of best fit would mean that it is where the graph originally starts because it is where the x-intercept is at 0. What the slope would mean in terms of the line would be that there is a predicted increase/decrease because it can be manipulated by the intercept.

Screenshot

amalialazonavarrete — 2024-03-16 05:30:40 UTC

Questions

amalialazonavarrete — 2024-03-16 06:02:10 UTC

The number of gallons from my household compares to the LSRL because when it comes to predicting my actual household water usage it isn't far off so it isn't a bad predictor. Also my actual water household usage is above the line of best fit and it is above the actual average of the amount of water usage.
My line of best fit and the LSRL are similar because they are both trying to find how close a data point is from the line and how well the data correlates from one to another. They are also different from each other because the line of best fit is a straight line that fits the best it can through the points while the LSRL shows the sum of the squares of the residuals.
The R2 shows how well the LSRL fits the data because with the number being closer to 1 it fits better because there is less distance between the data and more correlation, but with the number being further it shows how the data is farther from each other and the line doesn't really fit the graph.

Questions

amalialazonavarrete — 2024-03-16 06:15:55 UTC

Because the r-value is positive and so is the slope from the LSRL equation there is a moderate positive linear relationship between the number of people in a household and the amount of gallons of water used. Also with the R2 value being small and not that close to 1 I would say that there is not a close relationship between the number of people and amount of water used.
For confounding variables that could have led to residuals I would say it could have been location because maybe with a different household location there could a different price or less worry about water usage.

Questions

amalialazonavarrete — 2024-03-18 02:04:12 UTC

I think that the story that the data model is telling us is that because there are more people in a household there might be more conscience when using water because of the price. There are could be many reasons as to why water usage varies from household to household because the data isn't linear and so it isn't predictable. But if I had to sum up what the story would be about with the data it would be that when the amount of people in a household goes up so does the amount of water usage.

Questions

amalialazonavarrete — 2024-03-18 19:55:24 UTC

I think that this information would be good to predict the water usage for city or state as an estimate because of the large numbers and the way that the data comes from large populations.

I don't think that it would be good for larger households and school because even though they are larger numbers than most households, they wouldn't be large enough I think to use this information to predict water usage. I think that in order to make an accurate prediction we should use variables that are relatable or similar in data in order to get more accurate and realistic predictions.
The predicted calculations for our city's household water usage was 96,436.9. I think that the value was reasonable for the Roseville area because comparing it to the other data in both population and daily water usage it seems to follow a similar pattern and doesn't seem to be unrealistic.
For Roseville the water per week would be 675,058.7=(0.1327(151901)+76279.7(7)), per month was 20251762.2=(0.1327(151901)+76279.7(7)(30)), per year was 7,391,893,190.9=(0.1327(151901)+76279.7(7)(30)(365)).