BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Herwibawa
Calcite | Level 5

Hi,

I'm elementary user of SAS. I am working on agricultural research and my experiment was laid out in split plot design.

However, I'm facing a problem, in which several plants had died, even of plants in one of the treatment combination.

So, what should I do ?

Ok, I'll explain my data;

A1     B1     1      died

A1     B1     2      20

A1     B1     3      died

A1     B1     4      died

A1     B1     5      died

A1     B2     1      died

A1     B2     2      13

A1     B2     3      died

A1     B2     4      12

A1     B2     5      10

A1     B3     1      died

A1     B3     2      died

A1     B3     3      died

A1     B3     4      15

A1     B3     5      17

A2     B1     1      died

A2     B1     2      died

A2     B1     3      died

A2     B1     4      died

A2     B1     5      died

A2     B2     1      died

A2     B2     2      28

A2     B2     3      died

A2     B2     4      died

A2     B2     5      30

A2     B3     1      17

A2     B3     2      15

A2     B3     3      16

A2     B3     4      died

A2     B3     5      18

A3     B1     1      5

A3     B1     2      died

A3     B1     3      10

A3     B1     4      died

A3     B1     5      died

A3     B2     1      died

A3     B2     2      10

A3     B2     3      8

A3     B2     4      9

A3     B2     5      died

A3     B3     1      died

A3     B3     2      5

A3     B3     3      died

A3     B3     4      10

A3     B3     5      6


It would be appreciated if you could help me


A very big thank you for your help

1 ACCEPTED SOLUTION

Accepted Solutions
lvm
Rhodochrosite | Level 12 lvm
Rhodochrosite | Level 12

It all depends if the deaths are due to the treatment. If they are, then your dead plants cannot be considered missing at random, and you have problems. You can treat plant death as a binary variable (0: alive, 1: dead, one value for each record) and analyze the data as a generalized linear mixed model (check out the SAS User's Guide for GLIMMIX).

     model z = A|B / dist=bin link=logit;

if z is the binary variable. See the rest of the code in your other posts for this split plot.

This idea is ignoring the quantitative response variable where the plants did not die. If plant death is related to treatment, then you can definitely NOT analyze the continuous variables for the alive plants. If plant death is unrelated to the treatments (unlikely), you could analyze the rest of the data (you don't have to do anything special). You have a lot of dead plants, so you probably need better experimental methods -- statistics cannot rescue an experiment that should not be analyzed.

View solution in original post

6 REPLIES 6
Herwibawa
Calcite | Level 5

After I read several statistic book I knew that is missing value, but how to create the codes using SAS ? It would be appriciated if you could help me

RW9
Diamond | Level 26 RW9
Diamond | Level 26

Errrm, water them more?  Sorry, sarcasm aside, you have not specified any problem.  Try rephrasing it with test data (just a small amount) in the form of a datastep.  Provide example required output, and clarify any logical steps.

lvm
Rhodochrosite | Level 12 lvm
Rhodochrosite | Level 12

It all depends if the deaths are due to the treatment. If they are, then your dead plants cannot be considered missing at random, and you have problems. You can treat plant death as a binary variable (0: alive, 1: dead, one value for each record) and analyze the data as a generalized linear mixed model (check out the SAS User's Guide for GLIMMIX).

     model z = A|B / dist=bin link=logit;

if z is the binary variable. See the rest of the code in your other posts for this split plot.

This idea is ignoring the quantitative response variable where the plants did not die. If plant death is related to treatment, then you can definitely NOT analyze the continuous variables for the alive plants. If plant death is unrelated to the treatments (unlikely), you could analyze the rest of the data (you don't have to do anything special). You have a lot of dead plants, so you probably need better experimental methods -- statistics cannot rescue an experiment that should not be analyzed.

Rick_SAS
SAS Super FREQ

@lvm's response reminds me of a famous quote by R. Fisher: "To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of."

Herwibawa
Calcite | Level 5

A very big thank you @lvm, @RW9, and @Rick Wicklin

Herwibawa
Calcite | Level 5

My experiment arranged in Split Plot design. How to make a Charts/ Boxplots for Proportions of a Binary Variable ?

data Survival;

  input A $ B $ Block Y;  if Y=1 then 1='Survive'; else 0='Died';

datalines;

A1     B1     1      0

A1     B1     2      1

A1     B1     3      0

A1     B1     4      0

A1     B1     5      0

A1     B2     1      1

A1     B2     2      1

A1     B2     3      1

A1     B2     4      1

A1     B2     5      0

A1     B3     1      1

A1     B3     2      0

A1     B3     3      0

A1     B3     4      1

A1     B3     5      1

A2     B1     1      0

A2     B1     2      0

A2     B1     3      0

A2     B1     4      0

A2     B1     5      0

A2     B2     1      1

A2     B2     2      0

A2     B2     3      0

A2     B2     4      1

A2     B2     5      0

A2     B3     1      1

A2     B3     2      1

A2     B3     3      1

A2     B3     4      0

A2     B3     5      1

A3     B1     1      1

A3     B1     2      0

A3     B1     3      1

A3     B1     4      0

A3     B1     5      0

A3     B2     1      0

A3     B2     2      1

A3     B2     3      1

A3     B2     4      1

A3     B2     5      0

A3     B3     1      0

A3     B3     2      1

A3     B3     3      0

A3     B3     4      1

A3     B3     5      1

;

 

Is it true when it displayed in as shown below ? how to write the codes

 

 

original.jpg

 

 

Please help me it is so important for me...

sas-innovate-2024.png

Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.

Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.

 

Register now!

What is ANOVA?

ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 6 replies
  • 1811 views
  • 7 likes
  • 4 in conversation