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

Category and data collection Question

Post by HopePressler1 »

My daughter is applying glogerm to her hands and then washing her hands with different types of soap to see if one type is better at cleaning her hands than others.

What category would this be in for a science fair? Is it chemistry, medicine/health or consumer science or something else? She has to enter a category for registering.

Also after she applies the glogerm she will take a picture showing that her hands are entirely covered and the glogerm and another picture after washing her hands.

How will she convert that information into data? She thought about trying to designate different parts of the hand (palm, finger tips, back of hand, fingers) and then documenting the percentage of each area of the hand that still had the glogerm after washing them. I am concerned that it is not going to be exact to have a percentage. Looking at her hands and saying about whatever percent doesn’t seem valid to me. Do you think her data idea is valid or do you have another idea about how to document the findings?

Thank you
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 »

Hello Hope and science daughter,

This sounds like a very useful project!

As to which category, aspects of this project include all the categories you mention, but perhaps medicine/health would be most relevant, since your daughter is focused on which of the soaps cleans most thoroughly.

About measuring the effect of each soap, your daughter's idea about recording the effect separately for each "zone" of her hand sounds good. Maybe instead of trying to figure out how to measure percentages, perhaps record 1=presence, 0=clear for each zone. She might want to have more refined zones, perhaps divide each finger into 3 zones using the knuckle as a divider. Or maybe even have each knuckle be a zone. And, for the back of the hand, maybe have each knuckle be a zone.

Then she could make a table listing each zone, and record 1 or 0 for each zone. The table should have columns for each trial (3 or more if possible), and a table for each type of soap. Then she can compare the sum of the 1's for each trial for each soap. Having multiple trials will help her demonstrate how consistent her handwashing is for a given soap.

Also record the number of seconds for each "hand washing".

It also would help if she (you!) could mark her hand to show each zone, if she can find a marker that won't wash off. The photos are a great idea, and will make the project display board "eye-catching".

Let us know if these ideas help and if you have more questions. Good luck!
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Data Analysis Question

Post by HopePressler1 »

My daughter completed a science fair project asking the following questions - Which soap eliminates the most germs? and Do soaps that cost more eliminate more germs? She used GloGerm on her hands and washed her hands three times with 16 different soaps. Each time after she washed her hands she used a black light to see if glogerm remained on 22 different areas of her hands. If the area had glogerm remaining then she wrote down a 1 for that area. If no glogerm was remaining she wrote down a 0. She added up all of the numbers for the left hand and the right hand separately and averaged those two numbers. The lower the number was indicated the more effective the soap. Is there a data analysis test that can be completed to indicate if the differences are statistically different?
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 »

Hello Hope and science daughter,

This is an impressive science project! As a science fair judge, I suggest that your daughter start by making a plot of the data separately for each hand, showing the observed number of ones (or zeros if she prefers) for each of the 3 trials on the vertical axis for each of the 16 soaps, with the soaps on the horizontal axis. She could label the soaps by brand, and, since she also is asking "Do soaps that cost more eliminate more germs?", I suggest that she order the soaps by cost on the horizontal axis.

These two plots provide a visual summary that addresses several important aspects of the science project results: (a) Which soap eliminates the most germs? (i.e., has the hightest (or lowest, depending on how your daughter wants to phase it) counts) and (b) how consistent the counts are for a given hand for the 3 trials for each soap and (c) does the graph for the left hand look similar to the right hand.

If the soaps are ordered by cost (perhaps cost/ounce, if the soap bottles contained different amounts), then the visual summary should show if the counts increase/decrease by cost/ounce. She could also use Excel or Google Sheets to plot the regression line between counts and cost/ounce.

You asked for a data analysis to provide a statistical test for "which soap eliminates the most germs" ... the test that comes to mind is a chi-square goodness of fit, but that won't answer "which soap" so the more informative analysis is probably the (linear) regression looking at the counts by cost/ounce.

This is probably a more complicated answer than you expected. But this is a very nice science experiment which warrants a detailed analysis! Please ask more questions and consider posting a pdf of the graphs!
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 »

Hello Hope and Science Daughter,

It occurs to me that the judges might ask if there were certain areas on her hand that almost always were missed and did her handwashing get better during the course of the project ... If she recorded the data that could answer either of those questions, it would be great to include it. And, even if she didn't have the data, she might address those questions in her discussion about what she would do differently if she did a similar project in the future.

Best of luck at the science fair and let us know how it turned out!
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

This a pdf of her graphs.
Soap Effectiveness for Removing Glo Germ.pdf
(30.78 KiB) Downloaded 179 times
I am confused about what the linear regression would show. Her data shows that some soaps were more effective than others. It also shows that the price does not have a correlation with the effectiveness. How would the linear regression help explain that? I was thinking that she needed some sort of statistical analysis to show whether the differences she found were statistically different or not.

Would an ANOVA test or t test show if the differences she found were significant?

I like the additional idea of seeing if her hand washing was better over time. She has it all recorded in her notebook in the order that she washed her hands. She organized one graph by effectiveness and the other by price. I get her to look at whether the order of washing had a correlation to the effectiveness.

I will also have her look at areas that were always missed. I know she mentioned that her fingernail areas always had glogerm remaining.
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 »

Hello Hope and Science Daughter,

Thank you for posting the interesting graphs!

A couple of words of caution! You state "her data shows that some soaps were more effective" and "shows that the price does not have a correlation with the effectiveness"

Scientists and certainly data scientists use terms like "data shows" and "correlation" based on statistical analyses. When describing what the graphs "show", it is better to use vague terms like "the data suggest that some of the soaps removed more of the Glo-germ" and "the data suggest that there is no apparent association between price and effectiveness"

As for what linear regression would show: regression (linear or not) is used to evaluate the association between the counts of clean areas (or vice versa) and the cost per ounce. To calculate the regression, you would use the actual counts for each of the individual 3 trials and the cost per ounce. In a spreadsheet, you would have a column for the counts and a column for the cost per ounce, with one row for each of the 3 trials, for each of the 16 soaps. I suggest that you count the areas on each hand, so there would be the number of clean areas out of ?44 areas for each handwashing. It might seem obvious to have 6 counts, 2 hand X 3 trials, but the left hand and the right hand are "nested" within each trial, so the individual hand count is not independent! It is important to show the counts for each trial, since the variability of the counts between trials for each soap is important when looking at the variability between soaps.

A graph that shows the actual data for the regression analysis would place one symbol (a dot say) on the vertical axis for each trial for each of the 16 soaps, and put the soaps on the horizontal axis according to the cost per ounce. This graph would show the variability in counts and show the range of cost per ounce for the soaps.

You also ask about an Anova or t-test ... however, for both of those, one would wonder what the control comparison would be. In your daughter's experiment, it seems the important question is which soaps are removing the Glo_germ most effectively. Before thinking about a statistical test, I wonder about the ingredients in the soaps, or if liquid soaps differ from the bar or tablets. You mention that it was hard to remove the Glo-germ from her fingernail areas. Was it harder with the bar soaps? One way to look at this info would be to make a table so you can see if there are some obvious subgroups of soaps.

I realize that I am asking more questions than I am answering! Thank you for your patience! The best science projects result in data that elicits more questions! Your daughter is to be congratulated on her careful recording of the data and willingness to ask more questions! I look forward to seeing what she finds relative to "getting better" at washing! Thanks!
Madeline
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Madeline,
Thank you for all of this information. I have read through this with my daughter and she is creating a new graph like you suggested. I am attaching a sample graph with a few points of this data. Please tell me if this is what you were explaining. She discovered after comparing the ingredients in all of the soaps that the four soaps that the data suggested were more effective all had five of the same ingredients. None of the other soaps had all five of these ingredients. I thought that was really interesting.None of the soaps were effective in removing the Glo-germ from around her fingernails.

She used one bar soap, one foaming tablet soap, 2 foaming soaps, and the rest were liquid. I don't think grouping the soaps into those categories will show anything.

She did create a table with the ingredients in all of the soaps. I don't know what a graph would look like with that information. There was a very long list of ingredients from all of the soaps.

Hope and Sam
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 »

Hello Hope and Sam,
Wow! I'm impressed that you are both willing to dig into the data!

First, IMHO, Sam's finding about the fingernails may be the most important finding of this project! I'm not knowledgeable enough about the connection between Glo-germ residual and virus/bacteria residual, but Sam's study certainly should make us all keep our hands away from our face, even if we just washed our hands!!

Now, first to the ingredients: I'm not sure how best to make a graph, but maybe a table. By any chance, are those 5 "key" ingredients the first 5 in the list on the bottle? I've been told that the order of the ingredients goes from most to least. Or maybe look at if the label has "active ingredients"? So maybe make a table where the rows are the 16 soaps, and the columns are those 5 ingredients, and put a 1 for soap A in those columns if soap A contains them. Start with this and lets look at that ...

Now, the graph. Yes, showing the 3 trials with different colors is great. But (always a "but"!). Is the vertical axis showing the count of number of areas with Glo-germ residual? Then, maybe say something like that. As for the horizontal axis, I'm curious if the cost per ounce covers a wide range of values, or a narrow range. What you are showing is "ranking" the soaps, not the actual cost per ounce. So in a spreadsheet, for 1 trial for the left hand, the columns would be soap ID, hand, trial, # areas with residual GG, and cost/ounce. There would be 3 rows for each soap, 1 row for each trial. So we'd have 3 X 16 rows for the regression, which you expect will show no association, but that plot and the regression line would be good to show the judges.

Of course, as you know, the data may not show an association between residual Glo-germ and cost, but if that was an original hypothesis, it's probably worth getting a graph to show what did happen.

Let me know what you find for the ingredients and we might find a good way to graph that too!

Also, I'm curious what Sam found for whether the hand-washing got better with time?

Sam is to be congratulated on a well-designed experiment with such a careful lab notebook, plus asking all these questions!

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

Re: Category and data collection Question

Post by HopePressler1 »

Madeline,

Thank you again for your suggestions. Sam has enjoyed reading all of your responses and she is enjoying looking at the data in different ways. I am attaching the graph Sam made for the line regression analysis. Please let me know if you think that is the best way to display that data. She decided to eliminate the bar soap from the cost per ounce graph. The cost per ounce of bar soap is so different than liquid soap, and it does not seem like there is a way to document the true cost of the bar soap. Sam read several documents about the differences in the amount of liquid soap used per wash and the amount of bar soap used per wash. I tried to help her come up with a formula to state the true cost of the bar soap compared to liquid soap, but that seems like a whole different science fair project. Do you have any thoughts on that?

There are certainly differences in her hand washing each time, but there did not appear to be a trend of her hand washing getting better over time.

Only the soaps that are antibacterial list active ingredients. The others just list ingredients. When I had Sam look at the ingredients again she found that the ingredients that were common in the four soaps that removed the most Glo Germ were also common in three other soaps that did not remove as much Glo Germ so that is not as interesting a find as she had first thought. I am attaching the spreadsheet with that information. I was only able to attach it as a pdf. It might be hard to read. Please let me know your thoughts.
Attachments
Ingredients in Soaps - Sheet3.pdf
(63.44 KiB) Downloaded 132 times
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 »

Hello Hope and Sam,
It is great that Sam is enjoying looking at her data! IMHO, these are essential skills for doing science!

Thank you for showing the "line regression" graph! It does look like she has the data set up but I suggest changing a couple of the labels. First, it looks like she had the spreadsheet program make a "line graph" ... which just connected the "dots" for each trial. I wonder if the program has an option for "scatter plot"? (Tell me the name of the program and I can see if I can find the names of the options!)

A scatter plot should show the symbols for each trial for each of the 15 soaps (I agree to exclude the bar soap, but add a note about that to the list I suggest below!) Then hopefully she can get the program to calculate the regression, and plot the regression line (there should be only one line, not 3).

For the labels, I suggest the title of the graph say something like "Association between remaining Glo-germ and cost of soaps tested".

For the label for the vertical axis, I think she is showing the number of areas with Glo-germ remaining? If so, that is a good label! I think this is out of 22 areas? Then a nice detail would be to get the spreadsheet program to stop the vertical axis at 22!

I probably should have asked this before, but is the primary hypothesis about the cost of the soaps?

You are right, I can't read the table of ingredients ... maybe let's set that idea aside for the moment while she looks at the graphs ... I'm thinking the judges might also like to see another graph to show that the number of areas remaining did not show any association with time. I'm wondering how to select the color/shape for each symbol ... not sure she can show 15 different colors or symbols, but maybe she could use a number for each soap? Did she do all 3 trials for each soap and then go to the next soap? Maybe upload a photo of her logbook for this!

Another question the judges might ask is the volume of soap used for each washing? This highlights the quandary for how to handle the bar soap (and raises the question if there was any Glo-germ left on the bar of soap!)

I mentioned a "list" ... judges like to see a list of further questions, like what would she do differently, or what additional data she would record ... So top on this list would be how to deal with bars of soap vs liquid!

I have probably asked enough questions for this post! Let me know what makes sense and what doesn't! And be sure to tell me what spreadsheet program she is using so I can maybe help with how to get the regression!

Hope she is still enjoying this!!
Madeline
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Thank you for all the ideas.

This is a quick remake of the graph. She is using google sheets, but we also have Apple Numbers. There was an option for trend line in google sheets. I told her to click that option. Is that the same as the regression line? Sam and I don't understand what this shows though. Would it be better to average the three trials? There are so many dots.

If you compare this to the other graph some of the data is different. I was going over her data with her and we realized she had copied some of the prices incorrectly from one spreadsheet to another sheet. Its always so important to check your work. :-)

Yes, the vertical axis is the number of times Glo Germ remained on her hands.

I will have her create a graph with the soaps listed in the time order she washed her hands.

She does have a list of further questions and what she would like to do differently. While she was practicing going through her work her sister asked her the question about how much soap she used each time. She had used two pumps of soap, but her sister pointed out that the different pumps might give different amounts. She has written down that if she completed the project again she would use a measuring spoon to measure how much soap she used. This still does not cover the bar soap. I also pointed out that the foaming soap and liquid soap are different. There are always more things to think about.

She did was her hands three times consecutively with each soap. She also cleaned off all of the Glo Germ in between each time she washed her hands so there would not be a possible build up for the second and third trials. Now I am thinking it would have been better to wash once with each soap and then repeat the whole process three times. She should write that down as well.

We have included a picture of the raw data in her notebook. It is only letting me upload one picture for some reason.
IMG_9249.pdf
(1.08 MiB) Downloaded 186 times

Hope and Sam
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,

Nice graph, now we just need to ask google sheets for the overall regression! I will look at google sheets ... but, in about 20 minutes I am starting to judge (virtually) a regional science fair. I'll respond again later this evening.

In the meantime, be sure to keep that other sister around! What fun!

Madeline

Fortunately, one of the local science teachers gave me a tutorial on google sheets! And, for Sam and her sister's science projects for next year, you might look into a free online data analysis/graphing program called Tableau Public. I learned about that from the experts here at Science Buddies! When Sam learns that, then we start on another open-source graphics and statistics program! More later!!
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 »

Hello Hope and Sam,

I'm attaching a pdf that show the sample data and graph I created in google sheets. Sam's graph almost has all the steps. The differences are (1) I do not use the "trial" data at all, and (2) I use the actual numerical cost per ounce for the horizontal ("series") axis. Sam's graph looks like she was using the soap name as the horizontal axis. Notice that my graph has the cost evenly spread across the axis so that the several soaps that cost 0.21/ounce are all shown at 0.21 on the horizontal axis, whereas hers spreads the multiple instances of the same cost (e.g., 0.21) as though they had different values.

This graph shows what you and Sam expected to begin with ... the R^2 is essentially zero ... there is no apparent association between the number areas with Glo-germ remaining and the cost per ounce. But I think this is an important graph to include on the poster board!

I looked at the logbook info - that tells the judges lots of useful information! First is the careful detail defining the different areas, and then the very consistent results for the fingernails! I think this would also be good to include on the poster board, as well as a photo of the black-light lit hand, if that is possible? Also, is this page for just one hand?

Let me know if this makes sense!
Attachments
Sample Data for Glo-Germ project - with sample graph.pdf
(97.24 KiB) Downloaded 150 times
HopePressler1
Posts: 45
Joined: Sun Oct 07, 2018 12:20 pm
Occupation: Parent

Re: Category and data collection Question

Post by HopePressler1 »

Thank you. That makes so much more sense.

The picture and graph are for one hand.

This is the log book for the other hand. For some reason I couldn't attach both pictures yesterday.
IMG_9248.pdf
(1.41 MiB) Downloaded 159 times
Sam has not completed the graph for the other hand. I wanted to make sure she was getting the correct format with one hand first. I will go over this with her today.

Thank you again,
Hope
Locked

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