BIRYANI AND STATISTICS.
Statistical Hypothesis testing.
This testing is performed as a starting point of further investigation into Business problem statement with limited evidence about the world. It is an integral part of data science.
This case is specific to the vendor interest to standardize the approach. Because today Google reviews & stars pulled out from shadows and feature them front and center across search traffic. Considering that 45% of consumer won’t even look at your business with less then 4-stars.
Specific Case:
An Indian Restaurant in Montreal, Canada daily sale of biryani from online, walk-in and Dine-in is average count is 55 orders. From the Google review customers claim that quantity of biryani was reduce, from the original weight is 2.325lb. Due to this it believed that average sales per day are reducing.
Test Statistics:
Vendor taking the random sample size of 15 biryani pack from four different weeks. Sample standard deviation is 0.120 and alpha=0.10(margin of error). Determine the quantity is truly reduced.
Using Hypothesis testing, parameter and statistics to evaluate the claim.
Understand what is t-table, click here
t-statistic calculation
From the above normalization curve, critical value is -1.363, Hence t-calculated value is 0.193 which is higher than critical value and falls in the fail to reject Null Hypothesis region.
Decision: Fail to reject Ho Hypothesis, So we can say there is insufficient evidence to conclude quantity of biryani is reduced.
On other side, if it reject the Null Hypothesis then we have enough evidence to perform further evaluation using ML to meet business objectives to meet the success criteria.
Objective:
To Optimize the preparation process to reduce wastage proportional to the daily orders count or increase the quantity to satisfy the customer claims.
Success criteria: Increase sales count.
To Learn more about different statistical tests, Data Science and Machine Learning approach comment below.
