Is it possible to conduct an experiment without proper control where the effect of several doses of a particular nutrient on a particular crop will be evaluated? Is it possible to compare these doses with one another rather than comparing with a formal control using DMRT or turkey HSD? Again if we take two factors( two different nutrient sources and their respective level) is control is still needed as levels? For example if I take thee doses of nitrogen (100 kg N/ha, 150 kg N/ha and 200 kg N/ha) and three doses of potassium (50 kg k2O/ ha 100 kg k2O/ ha and 150kg k2O/ ha) in FRBD design; is it necessary to add another level as the control in nitrogen where no nitrogen will be applied?
If it involves in 'comparison', then the experiments need control groups.
Comparing only two groups can be possible using T-test, in this case, no control is necessary.
Where nutrients are involved, a control experiment is very important. It is the only way the outcome can reliably be attributed to the treatments. It is highly recommended.
If you are trying to establish the effects of treatments, the control in that case need not be about levels of the nutrients, but rather the 'usual practice'.
It depends on the objective of your plant nutrition research. If you are interested in seeing if e.g. a particular fertilizer should be recommended or not, then you need to compare that fertilizer to a control (note: a control could be no fertilizer or farmer's conventional practice).
If you are e.g. determining what rate/level of that fertilizer required to meet a nutrient demand, then a control is not 'absolutely' needed. If you still want to include the control, then you can treat the control a common level across you various factors or as a co-variate.
In short, it depends on the objective of your research.
It is greatly recommended to have a control group in plant nutrition experiments for the sake of comparison and coming up with appropriate recommendations for the nutrients under study. These can be negative or positive control groups. However, this may vary depending on the aims of the experiment.....
It is dependent on your objective. If the objective is to compare the two factors, then no need of control. If your interested in the effects of the two factors then you need to compare it with something thus the need of control
You have indicated you are testing several doses of nutrients. In addition you will use DMRT to separate the means. All these indicate you are conducting a factorial experiment. For such a case, simply include a control.
From the responses it's clear that you need to define to question better
Normally, you will need a control which normally will be a treatment without a level for better comparison of other treatments whether in the field or laboratory. However, depending on your objective you may decide to define what your control will be. On that basis you will be able to present your data in a way that it may be acceptable
It is necessary to compare other treatments with control and also within treatments. Control is considered the reference point or better still the natural state which informs of changes observed in other treatments. Therefore, is paramount to compare other treatments to it.
- Badges
- Science topic
- Similar topics
- Clinical Pharmacology
- Dosing