Return to Question

(note: I removed the original table because there were some mistakes in it. I preferred to use R annotation of Sal's answer.)

Hi everybody,

I got the following results for an experiment of fermentation to produce Butanol:

Butanol =c(11.7462,11.7904,11.9162,11.8732,11.8583,11.8677,11.8697,
11.7289,11.9296,11.7722,11.9813,11.9873,11.8058,11.8711,11.937,
11.8628,11.7786,11.7649,11.8459,11.9139,12.0537,12.1359,11.9949,
12.0752,11.9993,12.2802,12.2227,12.1274,12.1408,11.9896,12.1362,
12.265,12.1353,12.0812,12.511,12.1871,11.7881,12.1962,12.2482,12.189)

Strain = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2)

Enzyme = c("AH","AH","AH","AH","AH","AA","AA","AA","AA","AA",
"AL","AL","AL","AL","AL","B","B","B","B","B","C","C","C","C","C",
"AA","AA","AA","AA","AA","B","B","B","B","B","C","C","C","C","C")

myData = data.frame(Butanol, Strain, Enzyme)

myData$Strain = factor(myData$Strain)

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

Again, sorry for the newbie questions. I'm giving my best to learn it.

I got the following results for an experiment of fermentation to produce Butanol:

Butanol =c(11.7462,11.7904,11.9162,11.8732,11.8583,11.8677,11.8697,
11.7289,11.9296,11.7722,11.9813,11.9873,11.8058,11.8711,11.937,
11.8628,11.7786,11.7649,11.8459,11.9139,12.0537,12.1359,11.9949,
12.0752,11.9993,12.2802,12.2227,12.1274,12.1408,11.9896,12.1362,
12.265,12.1353,12.0812,12.511,12.1871,11.7881,12.1962,12.2482,12.189)

Strain = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2)

Enzyme = c("AH","AH","AH","AH","AH","AA","AA","AA","AA","AA",
"AL","AL","AL","AL","AL","B","B","B","B","B","C","C","C","C","C",
"AA","AA","AA","AA","AA","B","B","B","B","B","C","C","C","C","C")

myData = data.frame(Butanol, Strain, Enzyme)

myData$Strain = factor(myData$Strain)

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

I got the following results for an experiment of fermentation to produce Butanol (note: I removed the original table because there were some mistakes in it. I preferred to use R annotation of Sal's answer.)

Butanol =c(11.7462,11.7904,11.9162,11.8732,11.8583,11.8677,11.8697,
11.7289,11.9296,11.7722,11.9813,11.9873,11.8058,11.8711,11.937,
11.8628,11.7786,11.7649,11.8459,11.9139,12.0537,12.1359,11.9949,
12.0752,11.9993,12.2802,12.2227,12.1274,12.1408,11.9896,12.1362,
12.265,12.1353,12.0812,12.511,12.1871,11.7881,12.1962,12.2482,12.189)

Strain = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2)

Enzyme = c("AH","AH","AH","AH","AH","AA","AA","AA","AA","AA",
"AL","AL","AL","AL","AL","B","B","B","B","B","C","C","C","C","C",
"AA","AA","AA","AA","AA","B","B","B","B","B","C","C","C","C","C")

myData = data.frame(Butanol, Strain, Enzyme)

myData$Strain = factor(myData$Strain)

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

I loaded my results in R:

myData<-read.table('../Downloads/butanol.txt', dec='.', header=T)

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

Again, sorry for the newbie questions. I'm giving my best to learn it.

(note: I removed the original table because there were some mistakes in it. I preferred to use R annotation of Sal's answer.)

Hi everybody,

I got the following results for an experiment of fermentation to produce Butanol:

Butanol =c(11.7462,11.7904,11.9162,11.8732,11.8583,11.8677,11.8697,
11.7289,11.9296,11.7722,11.9813,11.9873,11.8058,11.8711,11.937,
11.8628,11.7786,11.7649,11.8459,11.9139,12.0537,12.1359,11.9949,
12.0752,11.9993,12.2802,12.2227,12.1274,12.1408,11.9896,12.1362,
12.265,12.1353,12.0812,12.511,12.1871,11.7881,12.1962,12.2482,12.189)

Strain = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2)

Enzyme = c("AH","AH","AH","AH","AH","AA","AA","AA","AA","AA",
"AL","AL","AL","AL","AL","B","B","B","B","B","C","C","C","C","C",
"AA","AA","AA","AA","AA","B","B","B","B","B","C","C","C","C","C")

myData = data.frame(Butanol, Strain, Enzyme)

myData$Strain = factor(myData$Strain)

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

Again, sorry for the newbie questions. I'm giving my best to learn it.

I got the following results for an experiment of fermentation to produce Butanol:

$$\begin{array}{c|c|c|} \text{Butanol} & \text{Strain} & \text{Enzyme} \\ \hline \ 11.7462 & 1 & AH \\ \hline \ 11.7904 & 1 & AH \\ \hline \ 11.9162 & 1 & AH \\ \hline \ 11.8732 & 1 & AH \\ \hline \ 11.8583 & 1 & AH \\ \hline \ 11.8677 & 1 & AA \\ \hline \ 11.8697 & 1 & AA \\ \hline \ 11.7289 & 1 & AA \\ \hline \ 11.9296 & 1 & AA \\ \hline \ 11.7722 & 1 & AA \\ \hline \ 11.9813 & 1 & AL \\ \hline \ 11.9873 & 1 & AL \\ \hline \ 11.8058 & 1 & AL \\ \hline \ 11.8711 & 1 & AL \\ \hline \ 11.9370 & 1 & AL \\ \hline \ 11.8628 & 1 & B \\ \hline \ 11.7786 & 1 & B \\ \hline \ 11.7649 & 1 & B \\ \hline \ 11.8459 & 1 & B \\ \hline \ 11.9139 & 1 & B \\ \hline \ 11.2802 & 2 & AA \\ \hline \ 12.2227 & 2 & AA \\ \hline \ 12.1274 & 2 & AA \\ \hline \ 12.1408 & 2 & AA \\ \hline \ 11.9896 & 2 & AA \\ \hline \ 12.1362 & 2 & B \\ \hline \ 12.2650 & 2 & B \\ \hline \ 12.1353 & 2 & B \\ \hline \ 12.0812 & 2 & B \\ \hline \ 12.5110 & 2 & B \\ \hline \ 12.0537 & 1 & C \\ \hline \ 12.1359 & 1 & C \\ \hline \ 11.9949 & 1 & C \\ \hline \ 12.0752 & 1 & C \\ \hline \ 11.9993 & 1 & C \\ \hline \ 12.1871 & 2 & C \\ \hline \ 12.7881 & 2 & C \\ \hline \ 12.1962 & 2 & C \\ \hline \ 12.2482 & 2 & C \\ \hline \ 12.1890 & 2 & C \\ \hline \end{array}$$

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

I loaded my results in R:

myData<-read.table('../Downloads/butanol.txt', dec='.', header=T)

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

Again, sorry for the newbie questions. I'm giving my best to learn it.

I got the following results for an experiment of fermentation to produce Butanol (note: I removed the original table because there were some mistakes in it. I preferred to use R annotation of Sal's answer.)

Butanol =c(11.7462,11.7904,11.9162,11.8732,11.8583,11.8677,11.8697,
11.7289,11.9296,11.7722,11.9813,11.9873,11.8058,11.8711,11.937,
11.8628,11.7786,11.7649,11.8459,11.9139,12.0537,12.1359,11.9949,
12.0752,11.9993,12.2802,12.2227,12.1274,12.1408,11.9896,12.1362,
12.265,12.1353,12.0812,12.511,12.1871,11.7881,12.1962,12.2482,12.189)

Strain = c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2)

Enzyme = c("AH","AH","AH","AH","AH","AA","AA","AA","AA","AA",
"AL","AL","AL","AL","AL","B","B","B","B","B","C","C","C","C","C",
"AA","AA","AA","AA","AA","B","B","B","B","B","C","C","C","C","C")

myData = data.frame(Butanol, Strain, Enzyme)

myData$Strain = factor(myData$Strain)

• The first column is the butanol produced (outcome) in the process;

• The second column is bacteria strain since I've conducted the experiment with two different strains;

• The third column is the enzyme to accelerate the butanol production, here I have five levels: AH (high dose of enzyme A), AA (average dose of enzyme A), AL (low dose of enzyme A), B (typical dose of enzyme B) and C (typical dose of enzyme C).

• 5 replicates per treatment

I loaded my results in R:

myData<-read.table('../Downloads/butanol.txt', dec='.', header=T)

Then, I ran the following code:

anova(aov(Butanol ~ Strain*Enzyme,myData))

Only interaction and Strain were significant, so I'm a little bit confused how to proceed.

• Am I using the correct SS Type (in my case is 1)? Firstly, I was in doubt if my data was unbalanced because I did not perform every factorial combination, so I would need to use SS Type 3.

• What should I do next? I think I should filter treatments with strain 1 and treatments with strain 2 and then, run anova for these subsets, right? Or should I run post-hoc directly for interaction?

Again, sorry for the newbie questions. I'm giving my best to learn it.