ANALISIS PENGARUH KUALITAS PRODUK, PROMOSI, DAN KUALITAS LAYANAN TERHADAP KEPUTUSAN PEMBELIAN (STUDI PADA TOKO EMAS KARUNIA DI CEPU JAWA TENGAH)

This study aims to determine the effect of product quality, promotion, and service quality, of purchasing decisions at Karunia Gold Jewellery Shop in Cepu- Jawa Tengah. The independent variable are product quality, promotion, and service quality affect the purchase decision as the dependent variable. The population of this study were customers of Karunia Gold Jewellery Shop in Cepu-Jawa Tengah, 75 responder was selected using nonprobability sampling technique. The purpose sampling is a technique to select sample only from responder who has using and buying product from Karunia Gold Jewellery Shop in Cepu-Jawa Tengah. The purpose of using purpose sampling technique is to make sure that the selected responder really know and familiar with Karunia Gold Jewellery Shop in Cepu-Jawa Tengah. Base on the result study, obtained the following equation of regression : Y1 = 0,265X1 + 0,269X2 + 0,267X3. Base on statistical analysis, indicator at this research have the character of valid and its variable have the character of reliable. On the classic assumption examining the free multicolonierity-regression model, it might not done heteroscedisty and normally distributed. Individually, the variables have greater influence is promotion variable with a regression coefficient 0,269, followed by service quality 0,267 and product quality 0,265. Hypothesis testing used t tests showing that the three independent variables are quality of product (X1), promotion (X2), and service quality (X3) which investigated proved positively and significantly affect the dependent variable is the purchase decision (Y). Then through the F test can be seen that the variables product quality, promotion, and service quality eligible to test the dependent purchasing decisions. Adjusted R Square explained figures that 45,7% of purchase decisions variation can be explained by the three independent varibles in regression equations. While the rest of 54,3%was explained by other variable outside of the three variables used in this stud