TWO FACTOR EXPERIMENTS
A common problem in research is investigating the effect of each of a number of variables, or factors, on some response ‘Y’. In earlier times factors were studied one at a time, with separate experiments devoted to each factor. Later, R. A. Fisher pointed out that important advantages are gained by combining the study of several factors in the same factorial experiment. In the factorial approach, the investigator compares all treatments that can be formed by combining the levels of the different factors. Factorial experimentation is highly efficient, because every observation supplies information about all the factors included in the experiment. Also factorial experimentation is a systematic method of investigating the relationships between the effects of different factors.
The individual treatments are called factors.
The treatment levels within a factor are called levels.
A 2 x 2 factor experiment with two factors and two levels for each factor is denoted as a 22 factorial experiment. An experiment with 'f ' factors at 't' levels is denoted as a 'ft' factorial experiment. If the number of levels in each treatments is different then the notation is tA x tB. For example, if factor A has 3 levels and factor B has 5 then it is a 3 x 5 factorial experiment. OPSTAT provides the analysis of most commonly used two factor experiments such as two factor CRD, Two factor RBD and Split Plot designs.The analysis of two factor experiments will be explained using the following data set:
Example of ANOVA for a 2x2 Factorial (CRD)
Data for the RCBD analysis of a 2 x 2 factorial design
|
|
Replicates |
|||
Character 1 |
1 |
2 |
3 |
4 |
|
Treatments combinations |
a0b0 |
12 |
15 |
14 |
13 |
a0b1 |
19 |
22 |
23 |
21 |
|
a1b0 |
29 |
27 |
33 |
30 |
|
a1b1 |
32 |
35 |
38 |
37 |
|
Character2 |
|||||
Treatments combinations
|
a0b0 |
12.5 | 15.7 | 13.4 | 14.2 |
a0b1 |
16.5 | 14.5 | 16.3 | 17.2 | |
a1b0 |
14.6 | 15.8 | 13.6 | 11.6 | |
a1b1 |
12.8 | 13.5 | 13.2 | 15.1 |
In this example we have two factor say A and B. Both A and B have two levels each i.e. 0 and 1.We have 2 x 2 = 4 treatment combinations. For analysis we have to arrange the data of these treatment combination as follows:
Data are arranged in data file or in text area of web page in such a way that all the replications of 1st level of first factor and 1st level of second factor (i.e. a0b0) should be entered in one line. The observations of replications must be space or tab delimited. Similarly for all other levels of 2nd factor corresponding to 1st level of 1st factor (i.e. a0b1)are entered. Now take the 2nd level of 1st factor (a1) and repeat all the levels of 2nd factor (b0 and b1) in similar manner. Note that each treatment combination must be entered in separate line. Moreover, if you want to analyse the data of many characters having same treatments then enter the data for second character just after the fist character without any line gap. The data file for above mentioned example will be looks likes
12 15 14 13
19 22 23 21
29 27 33 30
32 35 38 37
12.5 15.7 13.4 14.2
16.5 14.5 16.3 17.2
14.6 15.8 13.6 11.6
12.8 13.5 13.2 15.1
Procedure of Analysis
Click browse button to browse the data file and the press submit button if your data is present in a data file or type in the data in text area according to the data scheme explained above and then press the submit button just below the text area.
A web page will be displayed. Here you have to specify the levels of first factor, levels of second factor, Number of replications and Number of sets in the text boxes provided for these purposes.
Choose the appropriate design such Two factor CRD, Two Factor RBD, Split-Plot from Select Design Group box.
Choose the Transformation if needed.
Press the Analyze button to analyze your data.
The results will be displayed in separate web page.