i work with 3 variety of plant..these plant are submitted to Nacl test with 4 concentration (0.2.4.6) ..in every concentration i give plant 2 traitement with acid humiq. for exemple in concentration 2 g/l there are plant with acid humic H1 AND plant with H2 (2 groups) . i am confused how to choose the statistic test to know which is the best humic acid concentration and for which variety and which is the best concentration of Nacl ...IF SOMEONE can help me with detailled information ..i wil be grateful. (i attached a paper to explain better)
This seems to be a three way ANOVA. You have two between subjects factors (plant variety and acid type) and one within subject factor (concentration).
You can use Repeated Measures menu under General Linear Model in SPSS (as correct as I remember) to analyse your data.
In your dataset you must have a variable denoting plant variety, one variable for acid type and 4 variables for different concentrations.
You should put between subject factors to between subject box and repeateds (concentrations) to within subjects box.
Your between subject factors should be nominal and different concentrations should represent the value of your dependent variable at that concentration.
You could impose some models, including some interactions terms if you wish.
I do not have right now access to SPSS, but I can help you in case you have further questions.
Thank you for response ! what do you think about factorial model ? AND HOW TO KNOW IF my variable is dependant or independant ?
This is a factorial model with repeated measures.
Your outcome (response variable) is the dependent variable. The others are independent variables.
I disagree, it could be a model with repeated measures only if each indiviudal (each plant in this experiment) are measured across the time more than once, i.e. at the time 0, 15, 20 days for example.
This experiment is a 2x3x4 factorial design, the table attached nothing says about the number of replications.
In this case, if the response is measured just once at the end of the assay, then you can apply three way anova including all the interactions.
Don't forget test the ANOVA asumptions before run the ANOVA.
Yes you are right Luis, thanks for pointing out.
I implicitly assumed this, though it was not clear from the table. I should have asked to Maher about this.
If the doses are not given to same subjects then it would not become a within subject factor. So the design would become a factorial model.
Sorry to forget that ..all traitement are with 3 repelications.so i .can apply three way anova including all the interactions ? dancan test to compare means ? ..sorry of that but i am not very familar with statistics
Maher strictly speaking, replications are different from repeated measures.
Replications mean in each cell of the table, there are more than 1 experimental units.
Repeated measures are measurements taken from the same experimental unit on different conditions (like different times or doses/concentrations etc).
So do you have replications or repeats?
If you have replications then this means values in each cell are coming from different experimental units.
If you have repeats then this means values in each cell in a certain row are coming from the same experimental units.
If you have replications, your design is factorial with no repeated measurements.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA