Women’s Clothing E-Commerce Reviews LATEX

Introduction

As the Internet becomes more and more accessible, more people prefer to use online shopping rather than going to the shopping mall and choose the product. Thus, except for the description of the product in words and picture, the product review become more critical as people are making decision. Thus, it is not surprising that the topic of if the online review is trustworthy comes up. Through reading the researches and data, the results show that reviews tend to be more negative because people are more likely to complain the products

rather than praising them.

As part of my research, unfortunately, I can not find the data set exactly suit

for this purpose. But I found a data set of the Women’s Clothing E-Commerce Review, through this, I got data on age, the number of positive feedback, cloth ID, rating and class name which will be investigated the relationship in between or across them.

Statistics and Analysis

a. Single Variable Analysis

I. Age

According to the general data posted by the Market Pro Market, a little more than half (53%) of 18 to 29-years old and (47%) of 30 to 49 year old says they always or almost always read and write the online reviews when buying something for the first time. Fewer adults ages 50-64 (34%) or 65 and older 65 and older (23%) consistently engage in this type of online shopping behavior. Thus, these provides an important background information to understand the distribution of age while analyse the rating relationship to the age. So here is the graph of the age distribution:

There are around 20000 order reviews in this data set. Unlike the Market Pro states that, age from 35 to 40 are most people in data set rate the order and the population is around 4000. Age from 30 to 35 and the age from 40 to 45 have each around 2750 people rate the order. Therefore, the data suggests that this E-Commerce focus on more middle-age women rather than young women. While keep this in mind, it (age) will be an important indicator of the rating.

Like showing on the histogram, the calculation of the mean age of this data set is 43.21166119 and the median is 41 which means that most review are from middle age women. However, the standard deviation here is around 12.26 which means that there are a large range of people (in terms of age) in this data set, which means although most women are age around 40, there are people who are younger and older to engage

with this rating process.

II. Rating

This data set is central around the clothing and specifically by women. Thus, it is just interesting to look at the average of rating. So here is the graph:

This graph shows that most rating range from 5 to 5.5 (this is about 50% of the whole population). Most people rate their product relativity high with almost 75% rate it more than 4.

he calculation proves that more rate focus on 4 and 5 by the mean is 4.19 and the median is 5. With the standard deviation is around 1.11 means that the range of the rating was relatively small. As seeing the data of rating, it is a lot higher than I expected because most analysts webpage always title with ”online review are more toward negative.” I tried to find the reasons behind. According to Duncan Simester, only 1.5% of people ever write online reviews. And those who do leave online reviews are different from the average consumer. They are more likely to be married, have children, be younger, be poorer, have graduate degrees, buy things in unusual sizes and make returns. Thus, the population may be an indicator of higher rate number. Also, the different categories of the product can change the rating, too, as this data set only focus on the women clothing rating.

III. Positive Feedback Count

As the number of the positive feedback the product receive is another indicator of how well the product is seeing the distribution of the number of feedback can be the side evidence of the rating. The graph shows that most people give 0 feedback, which this population is around 50%, meaning that most people don’t give products any positive feedback. Almost 15% people give 2 positive feedback and few people give feedback more than 4.

The Calculation does proves it that the average number of the positive feedback people given is around 2.5. With the median is around 1 and standard deviation is around 5.7 which means that the range of the number of positive feedback is large, meaning some people give many positive feedback and some people don’t give any.

As we have to keep in mind that the population in this data set is women and the review is E-commerce, the way of measuring the number of feedback may not be the most accurate, since it should be reported by the customers. It is normal that people tend not to remember the exact number of recommendations to others. In addition, when the data set is recorded is important, since right now is during COVID-19, it affects our daily life in communication and lifestyle, which eventually affect the number of the positive feedback. Also, to be more accurate, it will be useful to collect the number of the negative feedback to compared the difference.

b. Bivariate Analysis

I.Two Quantitative Variables

The first graph calculates the relationship between the positive feedback and the rating. The assumption is that more positive feedback means higher rating and vise versa, because more you like it and recommend to others, the more likely that you like the product. Thus, giving the product a higher rate.

According to the graph above, visually, there isn’t a relationship can be found and as I calculated the r value which is −0.06583549352 which means a weak negative correlation. Thus, there suggests that there is hardly a correlation between the number of positive feedback and the rating.

The second graph explores the correlation between the age and rating, which as the previous single variable found that the main population is around women who age from 30-40. The assumption is that the older the people are, the lower the rating they will give.

As there is a weak positive correlation of 0.0257212292, it suggests that there is not really a relationship between the rating and age. The two above graphs show that both rating and age and the number of the positive feedback and rating don’t have any relationship.

It may because when analyzing, the data did not being sorted by the cloth ID and in general; or while rating is very subjective, the rating range varies a lot. If there are more information on each woman customer’s information on classes, race and etc (category by them), the result will more likely have a relationship.

II. Two Categorical Variables

The data compares the dresses and non-dresses (tops) with the relationship with the number of the division.With the assumption that there are more people buying the general petite dresses than those who buy the general dress. More people will buy tops than the dresses.

As the graph shows above there are more people buying the tops than the dresses and more people buy the general rather than general petite. For tops, there are around 67% tops are general rather than the general petite, while 60% dresses are general rather than the general petite.

The graph proves the observation above 66.8% order choose the general bottoms and similarly 66% orders choose the general tops. Thus, even though the total number of bottoms to tops varies a lot to 3307 and 9278, the proportion of choosing general and general petite remains the same.

III.Categorical Variable and a Quantitative Variable

This set of the data is looking at the relationship between the dress and non-dress and the rating, evaluating of the dress has a higher or lower rating compared to other products.

The amount of the dresses is around 5500 and most rating center around 5 to 5.5 which is about 54% of the rating. With the average rating of the dress being 4.156477438 and the average of the 4.191500677, it means that dress has a larger percentage has lower rating than other department/ products are.

Looking at the non-dress rating, most rating also center around 5-5.5 and there is approximately 76% of the 5-5.5 rating. Compared to the 54% of dress rating, other categories of the clothes has a higher percentage of having a 5-5.5 rating. The average of the non-dress data is around 4.204174348 which is higher than the

dresses, too。

3 Conclusion

There is around 20,000 data and the results show that the age and the rating doesn’t have much correlation and surprisingly the number of the positive feedback doesn’t have relationship with the rating either. But it is interesting to notice that most rating in this data set is relevantly high which consider to be the rage of 4 to 5. In addition, more order choose to general rather than general petite and more people buy tops compared to dresses. As this research, however, focuses on the women and according to the company research that “eighty-five percent (85%) of women consumers say they consider these e-commerce reviews extremely/very important when making a purchase, and savvy shoppers use them to ensure quality, to gather more insights when deciding between similar products, and to get the best deal.” Thus, analyzing the data of women review E-commerce is important for any fashion industry as they are improving the fiance. Along with dedicated online review sites, social media platforms now provide an opportunity for consumers to share their thoughts and experiences about the products and services they use. And this survey finds

that 39% of U.S. adults say they have shared their experiences or feelings about companies or products on social media sites like Facebook or Twitter. Some 55% of 18- to 29-year-olds have engaged in this behavior, as have half (50%) of those ages 30 to 49.