Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
comparing means
Message posted by Nathan Allen on April 23, 2000 at 12:00 AM (ET)
I am completing a ninth grade science project and need a simple way to analize the data. Model buildings were set up in compartments filled with sand, gravel, or bedrock and the frame was shaken. Rankings were recorded 1'st, 2'nd, or 3'rd according to when each building fell relative to the other two. After 50 trials I have averaged the rank orders to values between 0--3. How can I determine if these differences in average ranking are meaningful? My report is due May 1'st. Thanks.
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: comparing means
Message posted by Chad Allen on April 23, 2000 at 12:00 AM (ET)
Although I would first suggest changing the measured variable (the dependent variable) from a ranking to a time measurement--that way your using interval data instead of ordinal data. Performing this proceedure will allow a more prcise measurement and allow you to detect significnat differences with greater ease. However, either way you may use an 1-way ANOVA (w/3 levels of the independent measure). The F ratio you obtain will tell you if a significant difference exhists between two of the groups. If one does, perform a post hoc analysis to determine which two groups differ. I suggest either a Tukey's HSD or a Duncan's.Good Luck
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.