Completely Randomized Design
Suppose we want to determine whether there is a significant difference in the yield of three types of seed for cotton (A, B, C) based on planting seeds in 12 different plots of land. With a completely randomized design (CRD) we can randomly assign the seeds as follows:
Each seed type is assigned at random to 4 fields irrespective of the farm. The above represents one such random assignment. We can carry out the analysis for this design using One-way ANOVA.
Randomized Complete Block Design
See the following topics:
Hello,
I’ve been given an experiment with these things: 400 genotypes of soybean, with 3 replications and 3 treatments of different levels of salt stress with 1 control/check. I need to grow and examine soybean plants in a greenhouse at Nanjing Agricultural University, Nanjing, China with hydroponic system. My question is, whether CRD design will be suitable or RCBD? And also, can I get a rough design of it with proper guidelines? I’ve gone a lot of literature but couldn’t find proper way. Please respond to me as soon as possible. Keenly waiting!
Hello Kamran,
Which to use depends on many factors beyond those that you have specified in your comment. Usually, if you have sufficient time and resources you would use CRD. Otherwise, you might use RCBD.
Charles
Dear sir, I have a question. I am conducting an experiment of germination test by using two type of seed which is rice and lettuce seed. I need to conduct lab and greenhouse study for both type of seed. The seed will be coated using two method which are fiber and film coating. Which method should I use ya either CRD or RCBD? For the lab and greenhouse study.
The hypothesis of the study is the coating will enhance the germination for both type of seed. Plus, the germination characteristic that will be discussed are the length of roots, number of leaves and length of shoot.
If you have two types of seed and want to compare multiple characteristics, you need to use a multivariate test such as Hotelling’s T-square test. If, in addition, you want to compare two types of seed coating, then you could use MANOVA, the multivariate version of ANOVA.
All these topics are covered on the Real Statistics website.
Charles
This depends on (1) the hypothesis (or hypotheses) that you are trying to test and (2) any limitations that you have. Can you specify these?
Charles
Good day, I have a question,
I have 16 seedlings used to study the effect of giberellic acid on the rate of plant growth. I used 4 replicates and distributed samples in 4 treatments. The seedlings were placed in four different flasks, each containing a different hormone concentration in distilled water, 200 mL 0% GA, 200 mL 0.1% GA, 200mL 1% GA, 200 mL 10% GA. The change in stem height was recorded everyday for one week. What design should I use (CRD or RCBD) and how?
If Flask 1 contains 4 seedlings with 0% GA and Flask 2 contains 4 seedlings with .1% GA and Flask 3 contains 4 seedlings with 1% GA and Flask 4 contains 4 seedlings with 10% GA, you can use a CRD design comparing the four treatments at day 7 for example. If you want comparisons by day, things get more complicated and the test to use depends on the specific hypotheses you want to test. It is likely you will need a repeated measures design.
Charles
hello sir i just want to know the real life examples of CRD,RCBD and LSD with identify the treatments ,experimental units and response meaurement ?
please sir explain me real life examples
Isha,
I don’t have real examples to send you, but you can find many such examples by searching on the Internet. I found a lot of RCBD examples in agriculture when I prepared this part of the website a few years ago.
Charles
Isha, si deseas puedo ayudarte con varios ejemplos
miguel.hoyos811@gmail.com
Question plz, I want to sow four or five different variety of mustard (colza) with pea or broad bean now I do not know to use witch design (RCBD or CRD). my problem is that I don’t know to witch design I use plz help.
Generally, you would use CRD unless you have some economic or other constraints. Do you have any constraints?
Xharles
What is the similarities of rcbd and crd
Justine,
E.g., both approaches use ANOVA to determine whether groups are significantly different.
Charles
Hey Sir I work on Experiment on Crop with Drip and Furrow irrigation (2 Treatment) and there is 3 Replications. Please Suggest Which ANOVA is Applicable and how.
Whether to use ANOVA CRD or RCBD or something else depends on the constraints that you have and the hypothesis or hypotheses you want to test.
Charles
Hello sir,
I’m doing research in farm machinery and i have 2 treatments in five different row to row spacing of crop and no. of replication is 8, so please give me best suggestions for my research.
I have doing a research on avocado powder making.I use 3 different dryers for this process.Also use 4 different pretreatments before drying.Whixh one I shoul select as experimental design sir,CRD or RCBD
It depends on the hypothesis or hypotheses that you want to test and the constraints that you have.
Charles
Sir,
Good day!!!
I would lik to ask if is it possible to conduct split plot in RCBD it I have only 4 treatments for factor A, 4 for factor B and replicated 3x?
Hoping for your kind and response. Thank you
Hello Ricmar,
I haven’t yet included two-factor RCBD. In any case, I believe that you can conduct the analysis as described at
https://real-statistics.com/two-way-anova/three-factor-anova-without-replication/
The analysis which contains paired intersections should be used. Here factor A has 4 levels, factor B has 4 levels and factor, which is used for the blocks, has 3 levels. In this case, the SS and df values for AxC and BxC should be added to the Within (Error) SS and df terms and then removed from the analysis.
I will try to give an example on the website shortly.
Charles
Also, if im going to compare organic-based solution over chemical-based solution in pepper and lettuce what statistical design is applicable? thank you
You need to provide more details before I could answer your question.
Charles
I have 4 treatment, the effect of using arachis pintoi on the growth of mustard.
1) Control (No arachis pintoi)
2) Full Arachis pintoi (No nitrogen)
3) Full arachis pintoi and 1/2 Nitrogen
4) Full arachis pintoi and Full nitrogen
i will repeat this treatment for 3 times to get more replicates
wht should i use? rcbd or crd
Ahmad,
It seems like you have the requirements to use crd. If not you would use rcbd.
Charles
hello sir please assist me. I’m going to determine the best hydroponic solution for strawberry using 8 different hydroponic solutions,. What statistical design is applicable? is it CRD or RCBD? Appreciating your kindness and help sir!
It depends on the details of your experiment. Either approach can be used.
Charles
if im testing effect of pressure on seed placement uniformity of three seeds using a certain planter. what randomization design can i use and statistical test
This depends on the hypothesis that you are trying to test, the data that you can collect and the constraints that you have.
Charles
Hi sir i performed a field experiment to test the effect of artificially induced heat stress on wheat crop in field conditions. i used 205 wheat lines in three replication and one control. three replication are treated with terminal heat stress in field and control without treatment. i followed a RCBD design and repeated this experiment for two years.
now i am confuse how to get anova of this experiment. please suggest me how to get the ANOVA of my data
Hello Shabbir,
I don’t fully understand the scenario that you are describing.
Does 205 wheat lines mean that you have a Wheat factor with 205 levels (one for each wheat line)?
Since you have 3 replications, does this mean that you have a sample of size 205 x 3 = 615?
You say that you repeated this experiment for two years. How often did you repeat the experiment in these two years?
Charles
Is it possible to use CRD when the data is about pH level mitigation? The goal is to mitigate the acidity of soil and turn it to a higher pH level.
Yes, but it depends on the details.
Charles
I have a problem regarding the design i have to used. My research is about solar powered machine and I will test the machine at different time throughout the day what design is appropriate to my research?
Good morning guys,i need some serious help on my research design.I am working on a sorghum research and my greenhouse experiment requires 6 treatments and working on two genotypes(improved and local) with a pre requisite of 3 replications but i am extremely worked up on the layout and how it should be,HELP PLEASE!!!!
When you say you have 6 treatments, is this the 2 genotypes times the 3 replications?
If so, this looks like a simple t test (or Mann-Whitney of the assumptions for the t test are not met).
Charles
If my subjects are identical, do i use CRD or RCBD?
Jeff,
I need additional information in order to answer. Can you describe in more detail the scenario and the hypothesis you want to test?
Charles
Dear Sir,
How to interpret the comparison of means by Tukey’s provided by the tool?
In the 7th column, we get p-value for all treatment combinations. I that is < 0.05 then that combination is statistically significant. Is that right? Or is there some other method?
Hi Rohan,
Yes, any combination whose p-value is < .05 is significant (assuming a 5% significance level). The value of this test is that it automatically takes into account that multiple tests are being done (and so you don't need to adjust the significance level manually to account for family-wise error). Charles
Thank you
Show me brifly the similarity and diference b/n CRD and RCBD
This webpage attempts to do this. What needs clarification?
Charles
advanced biometery
I have a question about the CRD.
We have an experiment with 4 treatments and 4 plots per treatment with a CRD.
Per plot there is a certain number of measurements on individual trees, but (due to tree survival) this number differs per plot.
Is there a way the difference between treatments can be analyzed in MS excel?
Andries,
Can you provide more detail about the specific measurements you plan on conducting on individual trees?
Charles
We measure height and diameter per tree. Every plot contains normally 49 trees, but some trees have died or will die. We have 4 plots per treatment, and 4 treatments.
Andries,
Since you have two dependent variables, you could model this using some sort of multivariate design (perhaps a multivariate randomized complete block design). This is not supported on Real Statistics and I don’t know how to do this in Excel.
Charles
Question plz,. İf have a question in the form of table as above and it contains the data. How could i know wheather the solution will be in CRD or RCBD. The answer must be according to the direction of replications, my problem is i can’t differenciate plz help.
Mustafa,
Usually it is more cost-effective to conduct an RCBD rather than a CRD; otherwise you should conduct a CRD. See the following webpage for more details:
https://real-statistics.com/design-of-experiments/completely-randomized-design/randomized-complete-block-design/
Charles