Marketing strategies and organizational performance of Nigeria's selected tourist sites via multivariate regression technique

Introduction

Multivariate multiple regression (MMR) is employed to model the linear association between more than single explanatory variable (EV), but more than single response variable (RV). When MMR has more than one EV, it is said to be multiple. MMR is multivariate because there is more than one RV. This study looks at a scenario where the EV is five, while the RV is two. This is a situation of studying marketing strategies and organization performance of selected tourist sites in Nigeria, where the dependent variable is organizational performance comprising of two measures (market performance and brand image), whereas the independent variable is marketing strategies, comprising of five measures (Concentration Strategy, Distribution Strategy, Differentiation Strategy, Promotional Strategy and Pricing Strategy). Hence, the objectives of this study are;  to fit models for organizational performance (Market Performance and Brand Image) on the five marketing strategies (Concentration Strategy, Distribution Strategy, Differentiation Strategy, Promotional Strategy and Pricing Strategy) via Multivariate Multiple Linear Regression technique and  to test for the models as well as Multivariate Regression Parameters in order to draw meaningful conclusion.

Statement of Problem

Marketing strategies with organization performance has been employed in many research studies, especially in the areas of business related disciplines. Many related variables that were studied employed multiple regression, even when the dependent variable could accommodate more than one variable. It is as a result of the little or no knowledge of multivariate multiple regression by researchers in business area that led to this present study. The problem of not being able to work with more than one dependent variable will be address in this study, as two dependent variables will be jointly correlated with five independent variables.

Literature Review

In a work by Oguebu (2011) [7] , the response variables employed were average weight and bulk density of cheese balls, while the explanatory variables were oven temperature, moisture content before extrusion and moisture content after extrusion. Only oven temperature was significant in the model according to their result. A recommendation based on the result of the study suggested that larger dimensions of data set should be generated so as to incorporate more variables. Onyenawuli (2012) [8] did a work on a multivariate regression of measures of health versus socio-demographic characteristics of patients in Nigeria. The response variables employed were Blood Pressure, Body Mass Index (BMI), Pulse, while the explanatory variables were age, sex, marital Status of patients. The result showed that age, sex and marital status were significantly correlated to the dependent variable. Hence; the model did well as a predictive equation.

Ekezie et al (2013) [4] fitted a multivariate linear regression between dependent variables (Systolic Blood Pressure, Temperature, Height), being the vital signs and explanatory variables (Age, Sex of patients), being the social characteristics. SAS software package was employed to run the data for the study and the results revealed that the model was appropriate for joint relationship between dependent variables and explanatory variables. Further analysis of their study revealed that only age had impact on the vital signs of the patients. The study recommended that further study should increase the explanatory variables to at least four in order to compare results.

Model Techniques

Multiple linear regression has its basic model as:

Where E is the expectation, Var is the variance and C is the covariance Equation ( 2) is represented in matrix form as; where

and e has a dimension of ) 1 ( × n The unknown parameters are λ and σ 2 and the matrix Z has the vector

Multiple Regression in Multivariate Form

Examining an association for the RV;

, with m responses and sets of EV;

. Every assumed response has a model regression of its own, resulting in the following results as in Equation ( 5) The model for multivariate regression in linear form is

The regression model in linear form with ith response )

In the solution of a single response conforming, set

The results of collecting these estimates of univariate least squares are as follows:

The parameters' choice can take

, errors in the matrix become ZA V -. The cross-products matrix with the error sum of squares is

the generalized variance is minimized also by estimates λ ˆ of the least squares.

By employing the estimates λ ˆ of the least squares, the predicted values for the matrices and residuals can be formed as;

Data Analysis

The techniques discussed statistically in this study were employed to analyse the data collected for this study. The software R package was used to run the set of data for this study for easier computation. This implies that the models for the fitted multivariate regression are;

References

1. Cohn KV. Maximum Likelihood estimators in Multivariate Linear normal models. Journal of Multivariate Analysis 2000;33(4):182-201.
2. Echeta SO. Pricing/skimming Strategy and Performance of Obudu Mountain Resort in Nigeria. Unpublished Ph.D. Seminar Presented in the department of Hospitality and Tourism Management, Faculty of Business Administration, Imo State University, Owerri, Nigeria 2019.
3. Echeta SO. Promotional/Advertising Strategy and Performance of Obudu Mountain Resort in Nigeria. Unpublished Ph.D. Seminar Presented in the department of Hospitality and Tourism Management, Faculty of Business Administration, Imo State University, Owerri, Nigeria 2019.
4. Ekezie DD, Opara J, Anderson NC. Application of Multivariate Multiple Linear Regression Model on Vital Signs and Social Characteristics of Patients. International Journal of Engineering and Science 2013;2(6):40-56.
5. Ezeokeke C, Okoli C. A Comparison on Impact of HIV/AIDS Patients Characteristics on their Blood Pressure in Nigeria. European Journal of Engineering and Technology Research 2020;5(8):884-890.
6. Johnson RA, Wichern DW. Applied Multivariate Statistical Analysis. Prentice Hall, Englewood Cliffs, New Jersey 1992, 285 -328.
7. Oguebu CE. A Multivariate Linear Regression Study of confectionery products; A case study of Cheese Balls produced by Zubix International Company Ogidi between January 2010 to March 2011. Unpublished M.Sc. Thesis submitted to the department of Statistics, UNIZIK, Anambra State, Nigeria 2011.
8. Onyenawuli PA. A Multivariate Linear Regression Study on measures of health and socio-Demographic characteristics of patients. Unpublished M.Sc. Thesis submitted to the department of Statistics, Faculty of Science, Imo State University, Owerri, Nigeria 2012.