Data Science Padlet Project by Daniel Silva

This data used will be from a Google survey to all student in period 3 data science

danielsilva98 — 2025-02-27 20:09:08 UTC

Is household water usage and the number of people in a household correlated? Is this a casual or spurious relationship? Are there any confounding or mediating variables which help tell the story.

danielsilva98 — 2025-02-27 20:10:20 UTC

danielsilva98 — 2025-02-27 20:17:49 UTC

danielsilva98 — 2025-02-27 20:30:58 UTC

danielsilva98 — 2025-03-10 19:13:16 UTC

Looking at the Scatter plot we can see that for the most part it is pretty symmetrical. Although for this instance I think we would expect the graph to be more skewed left because having more people in a home we would expect more water to be used. A confounding variable could be how many bathrooms are in a house because some people might have to salvage the water they use.

danielsilva98 — 2025-03-10 19:16:44 UTC

Also looking at the last Scatter plot we can determine that per every 1 person that amount of water goes up 34.008 per gal by looking at the slope, Also looking at the y intercept we can determine that no point has gone lower than the y-intercept

danielsilva98 — 2025-03-10 19:21:01 UTC

My point compared to many is relatively low with how many people I have in my house, but with that I do have 2 babies under 3 and me and my sister are not home often so we don't use much water. We also do not drink from the sink or fridge only bottled water.

R2

danielsilva98 — 2025-03-10 19:23:54 UTC

Overall our R2 number is not bad, it did = .142 meaning our correlation is actually pretty strong but it could definitely still be better.

danielsilva98 — 2025-03-10 19:29:12 UTC

Looking at this graph comparing how many showers are taken each day by household, we can see of the start the correlation is very less strong with the r2 being .039 and the slope increase is not high. Honestly I expected the data to be skewed left again but this time it is actually almost opposite.

danielsilva98 — 2025-03-10 19:32:45 UTC

Now looking at this graph I can almost be certain that the strongest correlation would probably between water usage and people per household because we don't really track these things but the r2 decreased a lot from the last at .022 and the slope is only 3.81 and again the data is skewed right but the higher numbers are on the right side.