Difference in Proportions Test API
Find out if two proportions are statistically different.
> **Note:** This is depreciated.
Are two proportions statistically different? Suppose a user wants to compare two movies to determine if one movie has a significantly higher proportion of �likes� when compared to the other. With a large sample, there could be a statistically significant difference between the proportions 0.50 and 0.51. With a small sample, there may not be enough data to determine if these proportions are actually different.
This web service conducts a hypothesis test of the difference in two proportions based on user input of the number of successes and the total number of trials for the 2 comparison groups. In one possible scenario, this web service could be called from within a movie comparison app, telling the user whether one of the movies is really �liked� more often than the other, based on movie ratings.
Consumption of Web Service
This service accepts 4 arguments and does a hypothesis test of proportions.
The input arguments are:
Successes1 - Number of success events in sample 1.
Successes2 - Number of success events in sample 2.
Total1 - Size of sample 1.
Total2 - Size of sample 2.
The output of the service is the result of the hypothesis test along with the chi-square statistic, df, p-value, and proportion in sample 1/2 and confidence interval bounds.
There are multiple ways of consuming the service in an automated fashion (an example app is here).
Limitations
This is a very simple example for a test of difference in 2 proportions. As can be seen from the example code above, no error catching is implemented and the service assumes that all the variables are continuous.
While this web service could be consumed by users - potentially through a mobile app, website, or even on a local computer for example, the purpose of the web service is also to serve as an example of how Azure ML can be used to create web services on top of R code. With just a few lines of R code and clicks of a button within the Azure ML Studio, an experiment can be created with R code and published as a web service. The web service can then be published to the Azure Marketplace and consumed by users and devices across the world with no infrastructure set-up by the author of the web service.