Science Buddies: "Ask an Expert"

Title: The Effect of the Type of Metal on It's Amount of Disintegration in Hydrochloric Acid
I have qualitative data, which is ratio data to narrow it down.
I just need to know--
Which type of data analysis should I use for my project?

Hello laxgirl,

The appropriate analysis will depend on what data collect. Without some more detail about the experiment, it will be hard to offer specific guidance.

If you've already decided what it is you're going to measure, I'm sure someone here can offer you guidance on how to analyze and present the data effectively, but we'll need some more information on your project.

If you haven't yet decided what to measure, we can talk about what sort of measurements are likely to provide meaningful results.

Best,
Erik

I measured the amount of metal that disintegrated in grams on a triple-beam balance. I weighed the metal 2 days a week, for 4 weeks. Anything else I need to let you know about? Thanks Erik!

Hi,

I would start by plotting the data with a scatter plot. The amount of remaining metal would be the ordinate (y-axis) and time would be the x-axis.

What the plots look like should give you ideas about what to compare quantitatively. E.g., if the plots are straight lines, then comparing the slopes of the lines tells you the rate that the metals disolve.

Scientists really like to see plots of data; it tells them all your results in a single glance.

Okay, I am using mean for my central tendency and range for ny measure of variation. The purpose of this experiment was to find out how long it would take swallowed metal to disintegrate in someone's stomach. I have 4 metals: aluminum, copper, galvanized steel, and stainless steel. I performed 6 trials of each and took measurements a total of 8 times. Only one problem though, the aluminum disintegrated so much that it broke up into bits and pieces, causing me to be unable to measure this. Is there anything that I should do about that, keep in mind that I am done with my experiment and have disposed of the materials. Also, the stainless steel started to gain weught at the end, can you tell me why that happened? Another problem is that I accidentally threw away my project before I took any pictures, what do I do? Do I have to redo the experiment so that I have pictures because I was heard that I wouldn't recieve credit if I didn't have pictures. Because they would blame me of cheating. Is there anything else I should inform you about to make sure I do this as best as I can. Thanks Bruce!

Hi,

I assume the eight measurements were at different times; hence you have a time series for the data. I imagine a plot with all your data, with the axes I described. at each of the eight points, for each metal, you find a central tendency (median is nice because it dumps any outliers automaitcally) and draw your curves through those points.

The fact that the Al dissolved fastest is fine...that is what you were trying to measure.

As a annual science fair judge, I rarely question if the experiment was done or not. I worry more about the right balance of student and advisor work on the project. I don't think pictures, in my view, are an essential part. Graphs and data are much more important. Of course, I can't speak for any special rules your science fair may have.

The part about the stainless steel is very interesting. Be sure to include that in all your plots and all. I can guess at what happened (and so can you). Did you measure the pH of the acid throughout the experiment?

laxgirl wrote: Also, the stainless steel started to gain weught at the end, can you tell me why that happened?

This is really interesting. How much weight did it gain compared to the previous loss? Is it within your error of measurement?

I don't think I made a mistake but here are my data tables to observe.

Science Fair Data Tables


Initial Weight of Metals (grams) 10/13

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 24.6 2.2 12.7 6.0
2 26.0 2.3 12.2 4.9
3 25.9 2.1 12.4 6.0
4 24.4 2.2 12.5 5.1
5 23.0 2.3 12.4 5.1
6 24.6 2.1 13.1 4.9
Total 148.5 13.2 75.3 32.0


Visual:
Copper: no immediate reaction
Aluminum: little immediate reaction
Galvanized Steel: huge immediate
Stainless Steel: no immediate reaction



First Data Collection Weight of Metal (grams) 10/17

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 24.4 * 10.7 5.5
2 26.3 * 9.7 4.8
3 25.7 * 10.0 5.1
4 24.4 * 10.0 5.0
5 22.7 * 10.2 3.5
6 24.4 * 10.7 4.9
Total 147.9 ~ 61.3 28.8
Difference 0.6 ~ 14.0 3.2

*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: water is yellow-green; copper is cleaned but not dissolved
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is light yellow-green; metal is dark but has stayed intact
Stainless Steel: water is green to dark green; metal seems to have become darker and eroded slightly



Second Data Collection Weight of Metal (grams) 10/21

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 24.1 * 8.8 4.6
2 25.6 * 7.6 4.7
3 25.5 * 8.1 4.5
4 24.0 * 8.2 4.6
5 23.9 * 8.7 2.8
6 24.1 * 8.9 4.5
Total 147.2 ~ 50.3 25.7
Difference 0.7 ~ 11.0 3.1

*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: water is light green; copper is light but dark around the edges, showing erosion to start to take place
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is yellow-green; metal is dark but has stayed intact and the edges seem sliced
Stainless Steel: water is green to dark green; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath



Third Data Collection Weight of Metal (grams) 10/24

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 24.0 * 7.6 1.6
2 25.0 * 9.0 3.4
3 25.4 * 7.6 3.8
4 23.4 * 8.2 3.2
5 22.5 * 8.5 1.9
6 24.1 * 8.8 3.5
Total 144.4 ~ 49.7 17.4
Difference 2.8 ~ 0.6 8.3

*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: water is light green; copper is turning brown with spots, almost clean
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is yellow-green and has particles at the bottom of the jar; metal seems sliced
Stainless Steel: water is green to dark green; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath




Fourth Data Collection Weight of Metal (grams) 10/28

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 23.6 * 8.7 4.8
2 24.9 * 7.1 4.0
3 25.5 * 7.0 4.5
4 24.2 * 7.7 4.6
5 21.1 * 8.1 2.5
6 23.9 * 8.6 4.4
Total 142.7 ~ 47.2 24.8
Difference 1.7 ~ 2.7 -7.4

*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: water is light, light yellow-green to murky brownish; copper is dark brown
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is light yellow-green and has particles at the bottom of the jar; metal seems sliced and is a gray color
Stainless Steel: water is blue-green to dark green; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath



Fifth Data Collection Weight of Metal (grams) 10/31

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 24.3 * 8.6 4.9
2 26.9 * 7.1 4.5
3 25.7 * 7.4 4.9
4 25.3 * 7.7 4.7
5 21.5 * 8.4 2.9
6 24.4 * 8.6 4.7
Total 148.1 ~ 47.8 26.6
Difference -3.7 ~ 1.9 -1.8


*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: one jar is clear while others are dark yellow-green; copper is dark brown with spots to a mud-gray color
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is green and has particles at the bottom of the jar; metal seems sliced and is losing layers
Stainless Steel: water is green to dark green with small foam outline at the waters surface at first; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath




Sixth Data Collection Weight of Metal (grams) 11/4

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 22.8 * 8.1 4.4
2 24.7 * 6.9 4.2
3 24.9 * 7.1 4.4
4 24.3 * 7.4 4.4
5 20.0 * 7.6 2.5
6 22.5 * 7.5 4.1
Total 139.2 ~ 44.6 24.0
Difference 8.9 ~ 3.2 2.6


*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: one jar’s water is clear while others are murky brown; copper is both light and dark brown with spots to a mud-gray color
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is yellow-green and has particles at the bottom of the jar; metal seems sliced and is losing layers
Stainless Steel: water is green to dark green with small foam outline at the waters surface at first; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath




Seventh Data Collection Weight of Metal (grams) 11/7

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 22.0 * 7.4 4.5
2 23.4 * 6.1 4.3
3 24.2 * 6.3 4.5
4 24.0 * 7.8 4.5
5 19.1 * 7.5 4.3
6 21.5 * 7.5 4.1
Total 134.2 ~ 42.6 26.2
Difference 5.0 ~ 2.0 -2.2


*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: one jar’s water is light yellow-green while others are murkybrown; copper is almost a black to both light and dark brown with spots to a mud-gray color
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is yellow-green to an almost olive green and has particles at the bottom of the jar; metal seems sliced and is losing layers
Stainless Steel: water is green to dark green with small foam outline at the waters surface at first; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath



Eighth Data Collection Weight of Metal (grams) 11/11

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 20.7 * 7.9 4.7
2 21.9 * 6.6 4.1
3 23.7 * 7.1 4.6
4 22.2 * 8.4 4.5
5 17.9 * 8.2 4.2
6 20.3 * 7.8 4.3
Total 126.7 ~ 46.0 26.2
Difference 7.5 ~ 3.4 -0.2


*unable to weigh because metal was in bits and pieces
~ unable to calculate because of the inability to weigh the metal

Visual:
Copper: one jar’s water is light yellow-green while others are murky brown; copper is almost a black to both light and dark brown with spots to a mud-gray color
Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely
Galvanized Steel: water is yellow-green to an almost olive green and has particles at the bottom of the jar; metal seems sliced and is losing layers
Stainless Steel: water is green to dark green with small foam outline at the waters surface at first; metal seems to have become darker and the outermost layer looks as if it is almost rubbing off showing a lighter layer underneath


Total Data Lost Weight of Metal (grams)

Trial # Copper Aluminum Galvanized Steel Stainless Steel
1 3.9 ~ 4.8 1.3
2 4.1 ~ 5.6 0.8
3 2.2 ~ 5.3 1.4
4 2.2 ~ 4.1 0.6
5 5.1 ~ 4.2 0.9
6 1.6 ~ 5.3 0.6
Total 19.1 ~ 29.3 5.6
Average 3.2 ~ 4.9 0.9


~ unable to calculate because of the inability to weigh the metal


What do you think? Thanks!


Bruce,
I took eight measurements over 4 weeks, twice a week. But, I also had 6 trials of each of 4 metals.
Yes, but is it okay that I have no data in most of my tables and on any of my graphs for aluminum? Thanks Bruce!

Hi,

Well, you have to have the initial weight. If it dissolved before the first measurement in the acid, that could be put on the graph at the first time point with a down pointing arrow (plus written discussion in text) to show it went faster than your measurement interval.

OK laxgirl, now we got some data to play with!

First thing I notice, you're summing over the six trials. Try averaging them instead, and calculate the standard deviation (or, error of the average). Most scientific calculators have a statistics mode that makes this very easy, if you don't have access to one you may have one on your computer or maybe there's a web based calculator you can find.

When I average the six trials of stainless steel in your 5th, 6th, and 7th collections, I get
4.4 +- 0.8, 4.0 +- 0.8, and 4.4 +- 0.2, respectively. So like the sum, the average indeed rises again. And I see why: look at the 5th trial across all your data collections. Up to collection #6 it always came out much lower than the other five. From collection #7 on, this changed. Just from looking at those raw data, I would conclude something untypical happened here which we don't understand - this is sometimes called an 'outlier'. I personally would vote to discard the 5th trial for all your collections. You are in a comfortable position to do that because you have five others to base your analysis on (this is exactly the reason you want to take multiple measurements of the same kind).

Now look again at the numbers I posted. The error I quote is essentially the tolerance you can apply to the average. For example, 4.4 +- 0.8 tells you that 68% of your data points are somewhere between 3.6 and 5.2. Another way of looking at it is: the error is (0.8/4.4)=18% of your average. When you calculate the average and error for all your measurements, I'll bet most of your errors are going to be smaller than that, indicating that an 18% error is unusually large and therefore something is wrong. By discarding the 5th trial I'll bet your errors will in fact be smaller than 18%.

Luckily you have a clear-cut case which is easily argued. Plotting the weights from six trials of steel over time would also immediately show that one trial does not behave like the rest, and in fact behaves irregularly.

Hi laxgirl,

Thanks for posting all of your data - it looks like you have done a great job! If you ever decide to revisit this project in future, you can try and recover the mass of aluminum remaining by filtration. The residue can then be washed with distilled water to remove the acid, dried and then weighed. This would give you the mass you were missing in your data. In this instance, I wouldn't worry about it - Dr Weaver's suggestion of plotting this occurence is a great idea.

A couple of quick points on things I noticed whilst reading over your posting:

When the metals were placed in the acid, was the solution manipulated in any way? What was the temperature? Was it stirred or agitated, or just left to sit for the time the experiment was running? This will affect the wording that you have used in your observations. For example, you draw a conclusion that:

"Aluminum: water is yellow; aluminum is either eroded greatly or broken up profusely" (my emphasis added)

Do you really mean that the metal was eroded - or corroded? The mechanisms behind each are different. You may want to check the definitions and either correct your wording or make sure that your experimental design shows why it was one instead of the other. Does that make sense? This question may arise at the Science Fair!

Along the same lines, if the aluminum in this example *was* broken up - what was the mechanism behind this? Again, this goes back to your experimental design and whether or not the acidic solution was physically agitated or not.

Just some thoughts! And once again, well done on a great experiment!

Best wishes,
Geoff.