Posts: 4
Joined: Tue Dec 13, 2005 5:33 pm

how to analyze data

Postby elvisrocks » Thu Dec 15, 2005 7:54 am

My science research project is the effect of different amounts of gas and oil on the height and the health of the bean plants. The independent variable is the amount/type of liquid given to the plant. The levels of the independent variable are 1/2 tsp gas, 1 tsp gas, 1 tbsp gas, 1 1/2 tbsp gas, 1/2 tsp oil, 1 tsp oil, 1 tbsp oil, 1 1/2 tbsp oil and water(control). There are 6 trials for each level. There are two dependent variables, the height of the bean plants and the health of the bean plants. For recording the health of the plants I simply record "Healthy" or "Unhealthy".
The end data was as follows. The plants receiving gas all died by the end of the experiment. The experiment lasted 21 days. The plants receiving water had the highest mean height with 37.5 cm. The mean height for 1/2 tsp of oil was 29.5 cm. The mean height for 1 tsp of oil was 14.8 cm. The mean height for 1 tbsp of oil was 25.1 cm. The mean height for 1 1/2 tbsp of oil was 19.7 cm.
I am working on analyzing the data. Here is what I have so far. The height of the plant is quantitative ratio data. I am using mean as the measure of central tendency and range as the measure of variation. The health of the plant is qualitative nominal data. I am using mode as the measure of central tendency and frequency distribution as the measure of variation. Please tell me other ways I can furthur analyze the data. Should I use statistical graphs, stem and leaf plots,box plots, standard deviation or another way to analyze data?Thank you for your help.

Posts: 4
Joined: Wed Dec 14, 2005 2:27 pm


Postby jvsoccerplaya925 » Thu Dec 15, 2005 8:05 am

Hey Karla! It's Charlotte! Actually I don't have an answer....because to be honest....you know waaaaaaaaay more than me! Just saying hi though.

P.S: I <3 Jason!!

Posts: 31
Joined: Wed Sep 14, 2005 3:35 pm

Postby thetrans1ent » Sat Dec 17, 2005 1:20 pm

Box plots seem like a good choice for this experiment when it comes to comparing all 9 liquid combinations. For each you could also included a histogram.

Regarding using standard deviation: Use this only if your distribution of heights for each liquid combination is relatively symmetrical and not too skewed.

To measure spread you can also use IQR (interquartile range; Q3 minus Q1) instead of simply range (max minus min). This gets rid of outliers if you have any.

Return to “Preparing for the Science Fair”