# How to Analyze Rapfish: Multidimensional Scaling Part 3

How to Analyze Rapfish: After you are ready with the application and data, let’s start using the Multidimensional Scaling software. This tutorial seems long, but if you don’t follow it well, there will be errors that cause you to not be able to use this tool.

Okay, the picture below is the data that I have, the data that I will use as an example exercise in this third part. More clearly, how to edit this data is explained in the previous part 2.

The good and bad values are 0 and 5, which means that there are two kinds of scales used. There is a maximum value (value 5) that is considered good, and there is also a value of 5 that is considered bad. An example of a bad-value variable is a pest infestation population. We certainly agree that the higher the value of the pests that attack, the worse it is said to be.

Next, we open the application. You can read Part 1 again. Okay, I will try to briefly remind you how to open this MDS software. First open file 1_6, then select file-option-addins-update addins in the storage folder-select file 1_6. Then open the Excel file (usually the file name is RedSea). I use one folder for one worksheet. If there are 4 exercises, then I copy all the new folders and edit the redsea file. You may not do what I do, but in my experience, if I use add-ins outside the folder, a message will appear: “Bad Parameter”.

Ok, we open the file, and then we go to the RapScores sheet. We moved the data that we had before to this sheet.

Arrange your data as shown above. There is no need to be confused. I will explain it in detail here. Lines 2 to 4 I think you already know it. That is the scale-value data we have. Then in row 7 is the good value of all the variables above; to make it easier, the good row is shaded red. Row 8 is the bad value of all the variables, not shaded.

Row 9 (Up) is a row consisting of the first half of the variables with good values (shaded red) and the next half with bad values (not shaded). Just copy the good values from var0 to var4, and copy the bad values from var5 to var9. Conversely, row 10 (down) contains the first half of the variable bad and the next half of the variable good.

Then, starting from row 11 onward, we will create an anchor. This anchor will be in the form of a circle and limit all the attributes (in this case, kab A, B, and C), which will determine where the attribute is located. Starting from row 11, we count down as in column C.

This anchor unit can be created by copying the good value, then gradually going down the row one by one from the rightmost cell to the bad value. But note that the middle values (in columns H and I) are made parallel. Then, after there is only one cell with good value left, on row 18, the rightmost cell starts to be replaced with good value again. Keep doing that until the leftmost cell has a bad value and the rest has a good value.

The easiest way is to remember that the red shading contains good values and the unshaded ones contain bad values.

Next, in the add-ins tab in Excel, we click Rapfish. Then the Rapfish windows appear as follows:

The screenshot above shows the MDS application. We will change the value according to the data we have. To make it easier, I added a number to the image, and I will explain each one one by one according to the data we have so that it can be easily understood.

1. Is the number of fisheries; in this case, we fill in 3 according to the number of districts (A, B, and C).
2. We fill the districts in Excel starting from row 2, so we fill the value 2
3. The fisheries value is in column A, so we fill in A. That means we can set the Rapscore sheet fisheries in another column. Very well, just adjust it according to the conditions, and the Rapfish windows must be the same.
4. The value is 4, which contains Good, bad, Up, and Down.
5. Each reference value is in which row? Fill in each one. The good value is in rows 7, 8, Up 9, and Down 10.
6. We count how many rows of anchors we need to create in column C. In this exercise, we need 15. Fill in 15.
7. What row are the first anchors in? We fill in 11 because the first anchors are in row 11.
8. Number of attributes. Or what we usually know as the number of variables. We fill in 10 because we are using 10 variables. The number of attributes or variables should be at least two times the number of fisheries.
9. The variable starts in column D.
19. For the Monte Carlo value, how many replicates do we want? Here, I just filled in 5
20. Fill in the value 20.
21. I fill in the empty cells from left to right: 68, 96, 68, and 96.
22. This save and get function is to save our settings; if we don’t save the settings in this window, we will re-enter the numbers manually if we repeat the software run. For example, we will save it in fike param2.txt, then click save. And later, if we want to call the settings, we just write param2.txt below it and then click Get.

The next step is to click Run Rapfish.
Click Run Leveraging.
And click Run Montecarlo.
We see the results on each sheet.

The results are already visible. Our last task is to read the results of the MDS, which I will explain in the last part.

Continued: Multidimensional Scaling, Part 4