How to interpret The Rapfish Output. We have come to the final part of multidimensional scaling. This section will explain a little about the output produced by this tool. MDS will output values and charts. All of them reflect a good-to-bad scale with a certain level of sustainability.
Stress Value and Correlation Squared (RSQ)
The goodness of fit of the multidimensional scaling results is shown in the stress and squared correlation (RSQ) values in the RapAnalysis sheet.
The stress value in this exercise is 0.148141, with an RSQ value of 0.939709. A good model is indicated by the S-Stress value, which is smaller than 0.25, and the RSQ, which is close to 1. When viewed from the results of the stress and RSQ values, it can be said that the MDS results in this exercise are a good model and can represent the problem being discussed.
How to interpret The Rapfish Output: Index and Sustainability Status
This sheet also shows the rapfish ordination chart that illustrates the position of the three fisheries we processed.
The circle is the anchor or boundary. We can see that the higher the value obtained, the better. We can get the coordinates in the shaded part or in column H.
In the result, we get the following index data:
District A: 47.48616
District B: 53.15384
District C: 46.82169
The index category and sustainability status can be assessed with the following information:
Index value: 0.00–25.00 bad category: not sustainable
Index value: 25.01–50.00 less category: less sustainable
Index value: 50.01–75.00 moderate category: moderately sustainable
Index value 75.01–100.00: good category: very sustainable
So we can conclude that District A has a poor category, District B has a fair category, and District C has a poor category.
Multidimensional Scaling: Attribute influence/sensitivity analysis
Moving on to the next sheet, leverage Attributes. This sheet contains variables or attributes that strongly influence the assessment category of the three fisheries, or in this case, the insufficient or good assessment of the three districts. It can be said that if Cabin C wants to improve the index so that it can become a sufficient or good category, then this leverage variable can be used to accelerate the improvement of the index value.
In the figure above, it can be seen that variable 5 is the most influential variable in increasing the index compared to other variables. Then, followed below, are var4, var3, and var6. These four variables are said to be the leverage variables of the chili area category of the three districts that we processed in this exercise since parts 1, 2, and 3 before.
These four variables can be further processed to produce policies using PPA (Participatory Prospective Analysis), which generally and qualitatively describes the relationship between one variable and another so that it can be known which variables are better to do so as not to disturb the balance of other variables but directly affect the increase in the index. God willing, one day I will discuss PPA separately.
How to interpret The Rapfish Output: Monte Carlo
In the Monte Carlo sheet, we encounter repetition or repetition of the algorithm. This repetition is intended to assess whether the MDS output is sustainable or not. If described as a lever, this leverage determines the length of the lever. The longer the lever, the easier it is to increase the index value. while Monte Carlo is described as how strong the lever is. If the lever is long and strong and made of iron or steel, then it is certainly the right tool to increase the index value. And vice versa.
Then where is the Monte Carlo value? It is in column A for District A, column B for District B, and Column C for District C. If you use more than three fisheries, then the column is more than C (adjusting the number of fisheries).
In column A for district A, there are 10 rows. The first 5 rows are the Y coordinates, and the next 5 rows are the X coordinates on the Montecarlo graph. We take the Y value only and then find the average value. After that, we compare it with the MDS value.
The difference between Monetcarlo and MDS reflects their sustainability status. A difference value of <1 indicates that the status of the sustainability index at the confidence interval corresponding to the RSQ value is not much different. The small sustainability index between the two methods indicates that (1) the error in scoring each attribute is relatively small, (2) the variation in scoring each attribute is relatively small, (3) the analysis process carried out repeatedly is stable, and (4) missing data entry errors can be avoided.
We can conclude from the three districts that only district B has certainty or sustainability as a chili area. While kab A and C have a difference of more than 1, it is said that it is not sustainable or there is a data processing error, as previously explained.
I have finished my MDS material; hopefully it can be useful. This article does emphasize the steps you need to go through, not as a reference library. You can search for references with search engines and get quality scientific articles. If you use this tool and successfully make a review, I would love it if you didn’t mind sharing the news with me.
Thanks for reading.