Speaker2112
Harmless
Posts: 1
Registered: 3-9-2016
Member Is Offline
Mood: No Mood
|
|
Rye grown in Martian soil
The first data from my continuation of last years project that took me to states as an alternate. The first freshmen from my school to go. This
subject will be my 10th grade project as well. I am growing various crops, specifically rye, in soil that simulates the soil on mars. The data shown
is the mass of a 7cm piece of rye. Martian soil in red and plants grown in earth soil in blue. What do you guys think?
[Edited on 4-9-2016 by Speaker2112]
 
|
|
|
j_sum1
Administrator
Posts: 6321
Registered: 4-10-2014
Location: At home
Member Is Offline
Mood: Most of the ducks are in a row
|
|
Interesting concept for an investigation.
Round here, most will find the soil itself of more interest than the plants -- would you care to provide details of the martian soil that you are
simulating and how you have managed to produce something similar?
And what exactly is the composition of your control? What convinces you that your control will make a suitable comparison? If you have chosen a soil
that is good for growing rye then what does that tell you about your "martian soil"
You have produced some plant data but that graph is crap. It really tells the reader nothing. If you want to make a comparison between your two
tests of data you really need to construct two histograms on exactly the same scale. Then you will have something that can be compared. Excel does
not produce histograms except under great duress.
Is the difference between the two data sets significant?
You need to calculate the standard deviation for both data sets. Both the mean and standard deviation are easy in Excel.
Once you have done these calculations there is a process to find out if there is any statistical difference between the two data sets. The actual
calculations are not difficult but if you have not met it before you will need someone to guide you through the process and explain the reasoning
behind it.
IOW, this is a good start but some sophistication is needed in your data processing to make any sense of your results.
|
|
|
NEMO-Chemistry
International Hazard
Posts: 1559
Registered: 29-5-2016
Location: UK
Member Is Offline
Mood: No Mood
|
|
The data might mean more if you also used a clinostat to simulate the gravity or lack of on Mars.
Maybe Mars soil would grow Rye well for the conditions on that planet. Cool idea for an experiment but alor of variables to think of.
How did you make the Martian soil?
|
|
|
Metacelsus
International Hazard
Posts: 2539
Registered: 26-12-2012
Location: Boston, MA
Member Is Offline
Mood: Double, double, toil and trouble
|
|
I would actually advise against using a clinostat to attempt simulating lower gravity. It's better to test one variable at a time.
Also, by my calculations (F-test and T-test), there is no significant difference (P=0.05) between both the variances and the means. The actual
P-values are 0.37 and 0.49, respectively.
[Edited on 9-4-2016 by Metacelsus]
|
|
|
NEMO-Chemistry
International Hazard
Posts: 1559
Registered: 29-5-2016
Location: UK
Member Is Offline
Mood: No Mood
|
|
If there was a difference then surely that would only tell you that mars soil is/is not any good on earth?
What did you use for soil composition?
Was it sterile? Experiments like this are interesting.
Perchlorate in Martian soil!! I bet that made the pyro's sit up!
[Edited on 4-9-2016 by NEMO-Chemistry]
|
|
|
mayko
International Hazard
Posts: 1218
Registered: 17-1-2013
Location: Carrboro, NC
Member Is Offline
Mood: anomalous (Euclid class)
|
|
As others have mentioned, a histogram, boxplot and/or beeswarm/strip plot showing the two distributions would be most visually meaningful. For
quantitative analysis, you'll want a statistical test like the two-tailed independent t-test. Google around and you'll find helpful discussions like
this:
http://vassarstats.net/textbook/ch11pt1.html
You'll also probably want to make some snazzy, publication-quality graphics for your report and poster. I do general-purpose programming in Python,
which has decent statistical and graphing modules, but if you're interested in data analysis specifically, you might want to look into learning R,
which is built specifically for things like quick t-tests and dank graphs. Here are some useful links:
http://www.cookbook-r.com/Graphs/Plotting_distributions_(ggplot2)/
http://docs.ggplot2.org/0.9.3.1/geom_boxplot.html
http://www.cbs.dtu.dk/~eklund/beeswarm/
http://www.statmethods.net/stats/ttest.html
al-khemie is not a terrorist organization
"Chemicals, chemicals... I need chemicals!" - George Hayduke
"Wubbalubba dub-dub!" - Rick Sanchez
|
|
|
wg48
National Hazard
Posts: 821
Registered: 21-11-2015
Member Is Offline
Mood: No Mood
|
|
It would be more usual to use the weight of the whole plant in the analysis.
Depending how you selected the pieces and/or if you selected several pieces from each plant you may have introduce a bias or additional variance in to
the data.
If the reason for taking several pieces from each plant is to only increase the number of measurements or because your scales do not have the range to
measure a complete plant it would be better to calculate total mass of each plant then use those totals in your analysis.
|
|
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
