Amazon_Vine_Analysis

Overview

This analysis investigates whether or not there is a bias toward 5-star reviews among Amazon Vine members. To execute this analysis, the following process was followed:

1. Perform a basic ETL process utilizing an AWS relational database instsance connected to pgAdmin.
   • Extract the following dataset tool dataset: Amazon Tool Reviews
   • Transform the data into 4 DataFrames using PySpark:
     1. customers_table
     2. products_table
     3. review_id_table
     4. vine_table
   • Load the tables into a SQL database
2. Utilize Pandas to analyze the data to determine if there is any bias toward positve reviews among Vine members.

Results

• The total number of Vine reviews that were most helpful were 285.

• The percentage of 5-star reviews among Vine members was 57.19%, with a total of 163 5-star reviews.

• The total number of non-Vine reviews that were most helpful were 31,545.

• The percentage of 5-star reviews among non-Vine members was 46.33%, with a total of 14,614 5-star reviews.

Summary

• This analysis shows that there was a slight positivity bias for reviews among Amazon Vine members as compared to non-Amazon Vine members when reviewing tools.

• While there were far more non-Amazon Vine members reviewing tools (31,545) than Amazon Vine members (285), there is a slightly higher percentage of 5-star reviews among Vine members at 57.19% while non-Amazon Vine members gave 5-star reviews at a rate of 46.33.

• Further analysis could include some basic descriptive statistics of the two groups to see what the average star rating was for both Amazon Vine members and non-Amazon Vine members.
• Additionally, simple t-tests could be performed to compare the mean of the star rating among Amazon Vine members to the population mean as well as non-Amazon Vine members compare to the total population.