It takes a little effort to get MDS leverage in rapfish from R software. As we know, multidimensional scaling can use Excel add-ins or R software. Actually, SPSS can also use multidimensional scaling, commonly abbreviated as MDS, but I don’t understand whether the output already includes leverage and Monte Carlo or not.
Discussing the choice between Excel add-ins or R software is sometimes a dilemma. Excel add-ins offer complete results; not only graphs or output, but also all traces in the form of calculation numbers are left behind so that we can use them to the fullest. While the R application only outputs values and graphs, and even then not all graphs are made, including leverage and Monte Carlo. Not only that, the stress level assessment that shows the validity of the output is also questionable.
At first glance, it seems that Excel add-ins are superior. But it turns out that not everyone is lucky enough to have the appropriate settings and computers compatible with this add-on. I have encountered many complaints about always encountering “bad parameters” or failing to run macros. So that inevitably they use R software as an analysis tool.
I think talking about MDS output or sustainability value between R output and Excel add-ins is not too different. Because the output produced by both is clear, both in the form of graphs and numbers. However, it is different when talking about leverage.
Continuing my review of this software, I was curious about the leverage graph presented by Excel add-ins and R output. R only outputs the X and Y values of the leverage value, while readers or MDS users are more familiar with the bar graph to the right, where the longer the bar of a variable, the higher the influence on the level of sustainability.
So the question is: is the actual leverage that is usually displayed by the Excel add-ins output actually the X value or Y value issued by R?
This idea is actually not purely my own but was triggered by a reader’s question about creating a leverage bar in an R application.
So I tried to use both applications and process the same data to find out the answer. Indeed, social researchers, who are usually users, do not know the composition of the formula for determining leverage or the sustainability index. So this kind of problem is solved by experimentation.
Leverage in Rapfish From R Software
After running using the same data, I found the leverage data in the addins as shown above, and the R software output is also in the picture (table) above. I just made a chat bar from the X-axis and Y-axis in the R output and then matched it with the leverage results in Excel add-ins.
Here’s the picture I got:
In this figure, it can be seen that the order of the variables is close to the same or identical. The leverage variables in these two figures are capital, labor, land, and equipment.
How about using y values?
It appears in the figure above that there is no similarity between the Y values in the leverage output in the R software and the Excel add-ins output.
Then, what about the value?
The value itself is not exactly the same between the X axis in the R output and the Excel add-ins result. This is understandable because the value is used qualitatively. This means that the highest number indicates the leverage variable that has the most influence on sustainability, regardless of its value. It is rare to see this leverage value mentioned explicitly.
The only mention of value in MDS is when discussing the sustainability index. The conclusion obtained from this experiment is that the leverage bar can be taken from the x scores if using R software to process Multidimensional scaling.
Thank you.
Leave a Reply