Discussion - of hypothesis testing results (dissertation) - Essay Example

Logistic Regression Applicable in the same sense as the ordinary regression, logistic regression is applied to model a relationship or association of a dependent variable with one or more independent variables providing a foundation to find the model’s fit (Kocherlakota, 2010). In this case, the applicability of the logistic regression method is to allow for the estimation of an event’s probability of happening. For the current model, logistic regression was applied to test the fit of the model while at the same time predicting the probability of whether customers make the decision to make a purchase or not.

In this model, logistics regression was applied to test the probabilities that young people consider price, effect of brand image, design, trendiness of fast fashion cloths, and effect of advertising on their decision to make a purchase. Categorical variables were considered to show how various variables affect the predictability of the decision to purchase of not within the population. Based on gender, age, and education the logistic regression provides a platform of identifying fit within the model.

Chi Square The application of the Chi Square in this analysis is to identify the relationship or association of two cross-tabulated variables within a population (Kocherlakota, 2010). The null hypothesis in this case is considered to show that an association exists between the cross-tabulated variables of the population. The alternative hypothesis using the chi square analysis is used to refute the null and reinstate the hypothesis (Kocherlakota, 2010). In this case, the chi square method is applied to test the association between young people’s decisions of making a purchase with price, effect of brand image, effect of advertising, design, and/or trendiness of fast fashion cloths as influencing factors.

Log Likelihood Statistics The log likelihood statistics method was used in this analysis to show the fit between coefficient estimates and data (Kocherlakota, 2010). Since data is fixed, coefficients influence the likelihood that certain outcomes will take place. In the current case, the method was applied to show how different parameters influence the likelihood of young individuals to purchasing fast fashion or not. The parameters include price, design, brand image, trendiness, and advert effects on the decisions made.

Hypothesis Testing Discussion H1: As it mention in the literature by Jackson (2001), price is a major determinant of the decision to buy fast fashion. In addition, a conflicting factor is also identified. The cost of fast fashion cloths does not matter to young buyers provided that the fashion is trending. For current study, the hypothesis that price of fast fashion was a determining factor in making the decision to purchase fast fashion. Using the logistics regression, chi square, and log likelihood statistics results show that that there is an association of up to 75.

5% for the young buyers to consider price as a factor, 0.000 in chi square’s Asymp. Sig. (2-sided) test at both 1 and 5 percent significant supports the association of price with the decision to buy. These results validate the hypothesis.H2: Considering the brand image in hypothesis 2, it is shown that the -2 log likelihood statistics result decreases by 24.613 from the original model’s 105.306 to indicate a relationship between predictability of the decision to purchase or not. The chi square Asymp.

Sig (2-sided) result of 0.000 confirms the presence of an effect of brand image on purchasing decisions at both 1 and 5 percent significance level. These results validates the hypothesis. The hypothesis is also supported in the literature above by Guercini’s (2001) and Turconi’s (2010) arguments that brand image is a consideration for many young fast fashion buyers.H3: By applying the logistics regression, an ODD of 85.7% indicated that customers are not influenced by adverts. On the other hand, a chi square with Asymp. Sig. (2-sided) result of 0.

1730 invalidates the hypothesis and reinstates the null hypothesis that there are no association between the decisions to buy fast fashion clothes based and presence of adverts. The hypothesis in this case is rejected unlike Choi (2014); Tyler, Heeley & Bruce (2001) and Bruce & Daly (2006) arguments in this research literature that adverts influence the buyers’ decision to make a purchase. Among influencing factors capable of affecting the lack of association between adverts and the decision to buy fast fashion include tastes and preferences (Kazmi, and Batra, 2009).

Additionally, advertising agencies have been on record for providing wrong information through adverts thus leading customers to be skeptical and unable to make decisions to buy based on the adverts (Gaol, et al. 2014).H4: As it declared in the literature above the designs of fast fashion cloths influenced the decision to make a purchase (Cline, 2012). Additionally, Szmigin & Piacentini, (2014) argues that fast fashion companies thrive to provide customers with on-season cloths. Using the logistics result of 0.

371 for the decision not to purchase, 2-log likelihood effect of 5.341 from 106.804 for the model, and chi square result of 0.019 at 5% significance level support the association of design and young buyers’ decisions to purchase. Hypothesis 4 holds and the null hypothesis is rejected.H5: The logistic regression shows that there are 18.6 odds for the decision not to purchase. Out of 93 respondents, 5 decided to buy showing that the association between a decision to purchase and being trendy was weak.

Additionally, the chi square result that no effect to the decision to buy could be traced back to trendiness following considering Asymp. Sig (2-sided) outcome of 0.319 at 5% significant level. The validation of the null hypothesis contradicts with the literature of this study by Guercini (2001); Jackson (2001); Ko & Woodside (2013); and Okonkwo (2007) arguments that trends affect the decision of young consumers to purchase. Due to personal preferences, trendiness of fast fashion cloths does not affect the decision to buy as some are not appealing to the customers (Kim, Fiore, and Kim, 2013).

Additionally, due to differing communication channels, some customers do not keep up with the trends while others do not subscribe to the culture of adapting trends for fashion items (Schaefer, 1995).ReferencesBesanko, D., & Braeutigam, R. (2010). Microeconomics. Cengage Learning.Gaol, F. et al. (2014). Recent Trends in Social and Behavior Sciences; Proceedings of the International Congress on Interdisciplinary Behavior and Social Sciences. CRC Press.Havaldar. (2010). Business marketing: Text and Case, 3rd Ed.

Tata McGraw-Hill.I-Hsien, T. (2011). Social Networking Mining, Analysis, and Research trends: Techniques and applications. IGI Publishing.Kazmi, S, and Batra, S. (2009). Advertising and Sales Promotion. Mumbai: Excel Books India.Kim, E., Fiore, A., and Kim, H. (2013). Fashion Trends: Analysis and Forecasting. BergKocherlakota, N. (2010) The New Dynamic Public Finance: Quantitative Analysis: Methods and Results, Student edition. Princeton University Press; pp. 168-200Liljenwal, R. (2004). The power of point of purchase advertising: Marketing and Retail.

Point of Purchase Advertising Intl.Mandal, R. (2007). Microeconomic Theory. Atlantic Publishers & Dist.Moss, G. (2009). Gender Design and Marketing: How Gender drives our perception of design and Marketing. Humphries, UK: Gower Publishing.Rao, V. (2009). Handbook of Pricing Research in Marketing. Edward Elgar Publishing.Schaefer, H. (1995). International Economic Trend Analysis. Greenwood Publishing Group.