Hello every one am writing my Msc research proposal.
My experiment will test three feed sources (pulses) as a bee feed supplement. The trials will be set out "Completely Randomized Design. Each treatment group of bees will be provided with 150 g of the respective treatment diets. The DATA data to be collected will be amount of food consumed, sealed brood and bee bread areas and bee strength and honey yield. Data will be subjected to ANOVA using SPSS software (version 28) with proc mixed model of SAS, and Mean values will be separated using to Duncan’s multiple range test (DMRT) at p =0.05 (SAS Institute, 2012).
Is there anyone who advised me whether my data analysis methods are appropriate or not.
You don't give us enough details of your experiment design. you have 3 different treatments, but we don't know how many hives (=samples) you have in each treatment and we don't know if they are placed in one geographical point or more. If I was doing this experiment I would choose at least 10 hives for each treatment (because of the differences between each one) in each 3-4 different locations (at least). So you have 3 different levels that influence the differences/variations: the first (superior), that you have a control on it, is the "treatment". The second level, under the first, is the "geographical point". And the third level, the lower one, is the "hives". theses 2 last levels are random. I would choose nested ANOVA, which can give you the possibility to calculate the percentage of added differences by each level to the total variation: for 100% differences, you will find that TReatment gave you X%, the geographical point gave you Y% and the hives Z%.
For all explanations you can read in chapter 10 Sokal and Rohlf 1981
An other remark: in my knowledge 150g in a season per hive is not enough
third remark: the measures that you will prosecute (amount of food consumed, sealed brood and bee bread areas and bee strength and honey yield) will be done one time a year of more? if more you will have a fourth level (between treatment and geographical point) to the nested anova
good success
Itsic
Dear Dr. Itsic
Thank you very much for taking your precious time and provide your intellectual comments. Concerning hive numbers 16 including control (4 hives for each treatment). The location is in one place. The pollen substitutes that will provided to each colony is 150 gm per week. The measures (amount of feed, sealed brood and bee strength and honey yield) for one season. Considering the aforementioned information shall I go for nested anova ?
It looks like the statistical analysis methods that you have planned for your experiment are appropriate for testing the effects of different feed sources on bee performance.
Using a completely randomized design and analyzing the data with a one-way ANOVA, followed by a multiple range test (such as Duncan's test) to compare the means of the different treatment groups, will allow you to determine whether there are significant differences between the treatment groups in terms of the response variables that you have measured (e.g. food consumption, sealed brood and bee bread areas, bee strength, and honey yield).
Using proc mixed model in SAS to fit a mixed-effects model to your data may also be appropriate, depending on the specific design of your experiment and the nature of the response variables. This can help to account for any potential sources of variability or confounding factors that may affect the response variables.
- Badges
- Science topic