I have more than three treatment in my research. I am to use predictive analytical software, i.e. SPSS to compare this treatment. So I can't see Duncan multiple range rather i'm seeing post hoc
Duncan is one of many multiple comparisons tests(i.e. post hoc). Check the link below to see how to use Duncan in SPSS. Best wishes.
https://search.yahoo.com/yhs/search?p=How+to+get+Duncan+tests+in+SPSS&ei=UTF-8&hspart=mozilla&hsimp=yhs-003
The term 'post hoc' is a general term for the methods used to look at comparing means from an experiment. The comparisons should be only performed if they were pre-planned, and only if the F test for your particular factor is significant at your pre-determined level of alpha (typically 0.05). So a particular type of multiple comparison test may be applied, if and only if the factor was significant. Duncan's is a type of mutiple comparioson.
So Tukeys, also called Tukeys HSD, (Honest Significant Difference) is an example of a multiple comparison test. It is similar to a t test but takes into account the experiment-wise error.
In R after running a one-way anova with aov or lm and getting a model output, you would look at the F test for your factor or interest.
In R you can use the agricolae package, and then use the duncan.test command following on the model output (model) , where "factor" is the name of your factor.
duncan.test(model, "factor",console=TRUE)
The output will give the pairs of means and letters to determine whether they are differenec or not.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS