| dc.description.translatedabstract | The internet revolution has led many companies to the creation of electronic shops
and to the growth of electronic commerce. One of the e-commerce advantages is
the accessibility that provides. An online shop can be accessed 24/7, which is very
important for the consumers. his research, will focus in other factors influencing
the electronic commerce.
For this purpose data from ebay.co.uk were used. Data came from sales completed
in a period of 90 days and the seller was located in Cyprus. After the creation of
the database there was a categorization based on product type.
Some of the categories had a significant amount of data and were used for further
analysis. After the categorization, the final categories and the variables which were
suitable for analysis were chosen. These selected variables were product condition,
seller type, sales method, product price and shipping cost.
Data analysis was based in two analysis methods, hypothesis testing and binary
logistic regression. Both methods were applied separately in each category.
Looking at hypothesis testing for factors influencing the sale of a product were
searched based on the following:
Whether sold products percentage was higher when the products were new
or used (ProductCondition).
Whether sold products percentage was higher when the seller was business
or private (SellerType).
Whether sold products percentage was higher when the sale was listed as
bid or «buy it now» (BuyMethod).
Whether the consumers prefer buying products with low shipping cost
(ShippingCost).
At binary logistic regression product condition, seller type, sale method, price and
shipping cost were used as independent variables and sold (yes/no) was used as
dependent variable. Product condition, seller type, sale method and sold were
binary values so they were transformed into 1/0. Price and shipping cost were
continuous variables and they were used as is. After the data preparation, the
10 / 131
binary regression model for each category separately was exported, and found the
regression equation that predicts when an item can be sold and what factors are
important in order to get sold.
Finally in the last chapter, the results of this research are presented. The results are
different in each category and the conclusion is that there are different factors influencing sales in each category. Some of the factors though, are affecting most
of the categories. One of these factors is the sales method (Bid/Buy it Now). Based
on the research it can be said, that most consumers prefer buying from auction
because of the lower prices they might get. Similarly, it can be said that, when it
comes to antiques or high valued items, consumers prefer buying from business
than private sellers due to the authentication of items. | el_GR |
