Category and data collection Question

Ask questions about projects relating to: biology, biochemistry, genomics, microbiology, molecular biology, pharmacology/toxicology, zoology, human behavior, archeology, anthropology, political science, sociology, geology, environmental science, oceanography, seismology, weather, or atmosphere.

Moderators: AmyCowen, kgudger, bfinio, MadelineB, Moderators

HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Madeline,

You asked the question earlier if Sam's hand washing improved over the time of the investigation. Could she use a regression line to show whether or not it improved over time?

I attached a sample graph. This is adding the total of all three trials and comparing them to the order she tested them. If this would be a good way to show whether she improved over time I will get her to think it through and see if she can come up with the idea. I always want to guide them and not tell them, but I want to make sure that I am sending them in the right direction. Hope
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Madeline,

Another question - Should she also use a scatter plot and a regression line putting the soaps in order from least amount of Glo Germ remaining on her hands to most Glo Germ Remaining on her hands to show that the r squared value is high suggesting that some soaps are better at removing Glo Germ than others? She originally had used a bar graph. Hope
MadelineB
Moderator
Posts: 907
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist
Project Question: Interested in volunteering as an expert
Project Due Date: n/a
Project Status: Not applicable

Re: Category and data collection Question

Post by MadelineB »

Hi Hope and Sam,

I hope Sam is still having fun making graphs!

For the first graph, to look at the order of testing, it would be better if she treated each individual trial separately, using the counts of the areas with remaining glo-germ as the vertical variable (response, or dependent variable). The variable on the horizontal axis would then be the number corresponding to the order ... so the order variable for trial 1 for brand 1 is 1, trial 2 for brand 1 was the 2nd hand washing so the order number is 2, etc. Order will go from 1 to 3 times the number of soaps (is she excluding the bar soap from all analyses? are the foaming tablets more like a bar than liquid?)

The label for the horizontal axis is actually "Order" ... not brand. In Sam's case, she needs to add a column to the spreadsheet to show the order number and use that in the regression analysis, not the column that shows the brand name. It looks like google sheets might have created its own idea of order!

The basic idea of linear regression is "a regression model that estimates the relationship between one independent variable and one dependent variable using a straight line. Both variables should be quantitative."(from Google search). Here's a more detailed explanation: https://online.stat.psu.edu/stat462/node/91/

The same comments apply to the second graph - "brand" is not quantitative, it's not even numeric, so the only way google sheets could produce that "trendline" was to create a number, which doesn't have meaning in the "brand" sense. Showing the counts for each individual trial for each brand provides useful information - mainly the variability between counts for a particular brand, and the variability across brands. But it doesn't make sense to add a trendline.

If she is still having fun, there are 2 numeric measures for each brand that she might consider. This idea comes from that very valuable sister you mentioned: what if Sam could measure the volume or the weight of soap from one (or 2) plunges of the soap dispenser? If she has a scale (balance) to weigh that, she could put it on something like the lid of a cottage cheese container, and subtract the weight of the lid. An alternate, maybe more indirect, measure could be the (original) volume of soap in each brand, thinking that the larger bottles would have dispensers that dispense more. Hopefully, she has access to a balance, since she and her sister are obviously into doing science projects (yea!!!).

Thinking about Sam's project poster board ... if there's room, it would be fun to have a photo of all 16 bottles/bars of soap - not to illustrate the individual brands, but just to illustrate how busy Sam was doing this experiment!

Hope these suggestions continue to help!
Madeline
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Madeline,

That makes sense now about the soaps not being quantitative data. Is the bar graph the best way to show the soaps in order of least to most remaining Glo Germ on hands?

For the graph showing the order in which the soaps were tested is this the correct way to display it? This is a graph with the first 12 washings which is the first four soaps. I want to make sure I am understanding.
# with GG vs. Trial.pdf
(14.49 KiB) Downloaded 111 times
I think she should exclude the bar soap, but I am going to leave that up to her. We are still talking about that. I think the foaming tablets are the same as the other foaming soaps. It would have been better to test foaming soaps separately from liquid ones. I think that is the difference. Again I want her to think about all of those variables that didn't seem obvious at the beginning of the project. She could make separate graphs with foaming and liquid.

I like the idea about measuring the weight of the soap dispensed. That would be fun. We have a scale we have used for other projects like adding weight to rockets or cars and seeing how that affected their height and/or speed. Now I think we should have completed more data analysis then.

Would a graph with the weight of the soap have the weight on the horizontal axis and the Glo Germ remaining on the vertical axis? Would it be better to use the average amount remaining from all three trials for each soap compared to the weight? Is it correct to use weight instead of volume since she would be using a scale?

I was curious about seeing if her left hand and right hand had similar amounts of Glo Germ remaining. I had her make a graph of that. Her left seemed to have less than her right. I wonder if that is typical of right handed people. There is a whole other project in that one, but I don't think it is that important to this project. Although I do see why you said that the right and left data needed to be written separately.

Thank you again for the support. I have focussed so much on making sure my kids understand the scientific method when completing science projects, but the data analysis has not been our focus. Our projects will certainly have more data analysis in the future.

Hope
MadelineB
Moderator
Posts: 907
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist
Project Question: Interested in volunteering as an expert
Project Due Date: n/a
Project Status: Not applicable

Re: Category and data collection Question

Post by MadelineB »

Hi Hope and Sam,

I will try to answer your questions in order ...

1. Is the bar graph the best way to show the soaps in order of least to most remaining Glo Germ on hands?

A scatter plot, showing the # areas with GG for each trial for each soap, is more informative than a bar graph. The "information" about the effect of the soap is in the # areas for each individual trial, which is best shown with a symbol at the relevant height on the vertical axis

2. The format of the figure you uploaded looks good. If possible, it would be nice to have the color or the symbol of each point change for each brand of soap, I'm not sure how to do that in google sheets, so don't worry!

3. About foaming soaps and weight and volume: If you have a small graduated cylinder to measure the volume produced by 1 push of the pump dispenser, that would be great. I'm curious if the "foam" is produced by pumping the dispenser? Or does the bottle contain the foam? Is the purpose of the foaming to make it easier to get a could lather when washing? I wonder if a given volume of foam might weigh less than the same volume of the liquid soap? She could calculate the weight (milligrams) per volume (milliliters), which would have the units "mg/mL" ...

Interesting about the scientific method! The standard project guide usually talks about 1 dependent variable, 1 (or at most 2) independent variables, 1 control variable. In real life, setting up an experiment means thinking careful ahead of time about the "uncontrolled" control variables, such as the time for each washing, the order of the trials and brands, amount dispensed, ingredients, foaming, bars, tablets, even left hand or right hand, etc! You're right to emphasize the method, and it turns out, recording all these "uncontrolled" variables is actually an integral part of a science project or a lab experiment in real life! It's great that you are coaching your daughters! It also sounds like you must have a well-stocked home lab! That really helps!

Thanks for being interested in my questions! I look forward to your's and Sam's next questions!! As well as seeing next year's projects!!

Madeline
Locked

Return to “Grades 6-8: Life, Earth, and Social Sciences”