I have an experimental design where 5 treatments ([cadmium], mg/L) were applied to different plants (n = 6 plants per treatment, total n = 30). However, each treatment was NO applied sequentially to the same individual over time, BUT the response variables (photosynthesis and conductance) were measured three times per treatment and plant (time 0, time 5 and time 10, in days). I measured photosynthesis and conductance in leaves without replication (one measure per plant). I need to know whether there is a significant effect of the treatment (fixed factor, n = 5 levels) and time on photosynthesis and conductance. I suppose that plant should be added to the model as random factor (plant nested within treatment?) due to an inherent variation among subjects. I would use a repeated-measures design because more than one measurement of the same response variable is taken from each individual, but in this case it is not due to five different treatments applied sequentially, but rather to the effect of time. The effect of each treatment was measured three times. Treatments were independent, as they were used in different plants. I am interested in knowing such differences among treatments and, if possible, the interaction between treatment x time. I do not know how to use this more complex ANOVA, if it should be a repeated-measures design or a randomised block design. I even thought in running the simplest model, using treatment (fixed factor), plant (random factor nested within treament), time (covariable) and the interaction between treatment x time. Please, find attached an excel sheet with the data, where Photo1, Photo2 and Photo3, and Cond1, Cond2 and Cond3 are the measurements for each response variable (photosynthesis and conductance, respectively) over time.
Dear Alberto
I would suggest my articles to you. I hope it can be useful for you.
Mehrdad Zarafshar, Moslem Akbarinia, Hossein Askari, Seyed Mohsen Hosseini, Mehdi Rahaie, Daniel Struve and Gustavo Gabriel Striker. 2014. Morphological, physiological and biochemical responses to soil water deficit in seedlings of three populations of wild pear (Pyrus boisseriana). BASE journal. Volume 18(3): 353-366.
I agree with Repeated Measure ANOVA as you can find the effects of your treatments as between effects and effects of time as within effect.
Good luck
Dear Mehrdad,
Sure, it is much clearer now!! Nice, way to go! :)
Thank you very much for your reply and the paper, it was really of paramount importance to solve the matter.
Best regards.
Yes, this is a straightforward mixed ANOVA. Treatment is a between-subjects (between-plants) factor and time is a within-subjects factor. You could do mixed ANOVA or you could do linear mixed effects.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA