Market Research Project Report: Footwear Industry Differentiation
Market Research Project Report: Footwear Industry Differentiation
Market Research Project Report: Footwear Industry Differentiation
Submitted by: Mangesh Patil | Riddhi Biswas | Mahtaab Kajla | Makwana Ravindra Govindbhai Parimal Kumar Shivendu | Sachin Kumar | Sudipta Mandal
TABLE OF CONTENTS
Introduction, problem statement and objectives ..................................................................................... 5 Introduction ........................................................................................................................................ 5 Problem Statement and Objective of the study.................................................................................... 6 Rationale of the project ................................................................................................................... 6 Objective ......................................................................................................................................... 6 Market research Problem ................................................................................................................ 6 Approach to Problem & Scope of Study ........................................................................................... 6 Recent trends & developments ............................................................................................................... 6 Objectives and detailed methodology ..................................................................................................... 7 Literature Review ................................................................................................................................ 7 Research Design .................................................................................................................................. 8 Preparatory Research ...................................................................................................................... 8 Secondary Research ......................................................................................................................... 8 Information needs ................................................................................................................................... 9 Data Collection from Secondary resources ....................................................................................... 9 Data collection from Primary resources ........................................................................................... 9 Focus Group Discussion ......................................................................................................................... 10 Particulars of Focused Group Interview ............................................................................................. 10 Discussion Questions and Answer Summary ...................................................................................... 11 Findings ............................................................................................................................................. 12 Descriptive research .......................................................................................................................... 12 Scaling techniques ......................................................................................................................... 12 Questionnaire development and Pretesting ....................................................................................... 12 Pretesting ...................................................................................................................................... 13 Sampling Technique .......................................................................................................................... 13 Fieldwork .......................................................................................................................................... 13 Data analysis Procedure ........................................................................................................................ 13 Preprocessing the data ...................................................................................................................... 13 Editing ........................................................................................................................................... 13 Coding ........................................................................................................................................... 14 Analysis of data ..................................................................................................................................... 15 Cluster Analysis ..................................................................................................................................... 15 Chapter: Table of Contents Methodology and plan ...................................................................................................................... 13
Advanced Methods of Marketing Research National Footwear Brands ................................................................................................................. 15 Variables used ............................................................................................................................... 15 Hierarchical Clustering ................................................................................................................... 15 Defining the number of clusters: elbow rule .................................................................................. 17 Defining the number of clusters: Dendrogram ............................................................................... 17 ANOVA .......................................................................................................................................... 18 Post Hoc - Scheffe .......................................................................................................................... 21 Cluster Profiling ............................................................................................................................. 23 Foreign Footwear Brands ................................................................................................................... 23 Variables used ............................................................................................................................... 23 Hierarchical Clustering ................................................................................................................... 24 DEFINING THE NUMBER OF CLUSTERS: ELBOW RULE ..................................................................... 25 Defining the number of clusters: Dendrogram ............................................................................... 26 K-Means Clustering ........................................................................................................................ 27 Multiple Regression Analysis ................................................................................................................. 29 National Brands Footwear ................................................................................................................. 29 Dependent Variable ....................................................................................................................... 29 Independent variables ................................................................................................................... 29 Estimated multiple regression equation ......................................................................................... 30 Foreign Brands Footwear ................................................................................................................... 31 Dependent Variable ....................................................................................................................... 31 Independent variables ................................................................................................................... 31 Estimated multiple regression equation ......................................................................................... 32 Factor Analysis ...................................................................................................................................... 33 National Footwear Brands ................................................................................................................. 34 Variables considered...................................................................................................................... 34 Foreign Footwear Brands ................................................................................................................... 35 Variables considered...................................................................................................................... 35 Structure Equation Modeling ................................................................................................................. 36 Foreign Footwear Brands ................................................................................................................... 38 References ............................................................................................................................................ 39 Textbook References ..................................................................................................................... 39 Web References ............................................................................................................................ 39 Database References ..................................................................................................................... 39 Chapter: Table of Contents National Footwear Brand ................................................................................................................... 37
Forward
2 a.m., 25th August 2010. Just a final review and we can submit it, announced Sachin as we all compiled our research gathered in bits and pieces in the past two months on this project. The work has been a combined effort of our team and all of us have tried our best to gather the most accurate data and information. Analysis has been based on the available facts and data gathered from various sources, research as well as our intuitive understanding of the various aspects of the footwear industry. We tried to incorporate the elements of our learning in our own approach of working as a team so as to eliminate inefficiencies and bank upon the competencies of each individual member while allowing everyone to explore the marches of their comprehension and creativity. This project has been a wholesome learning experience for us and we would be glad to extend the learning process by welcoming criticism and suggestions on our work.
AMMR Group 1
Chapter: Forward
Acknowledgement
It gives us immense pleasure to acknowledge all those who have given their time and energy to supply all valuable facts and opinions that has helped us in bringing out this project to fruition. We would like to express our gratitude and respectful thanks to Prof. Atanu Adhikari for constantly supporting and guiding us in achieving the prescribed objectives of the research. Finally we would like to express our thanks to all our respondents and friends who were instrumental in the successful competition of our project.
AMMR Group 1
Chapter: Acknowledgement
brands were more popular in urban India whereas Indian brands such as Liberty, Relaxo, Lakhani were more popular in semi-urban and rural India. Price plays a big role in the market penetration as Indian brands were as much as 50%-60% cheaper than international brands. Changing purchase behavior of women where they preferred to often change their footwear according to their clothing and enjoy browsing city markets to purchase footwear from footpaths and smaller stores it is cheaper. Women also like to buy the latest footwear and the smaller shops with private labels provided that opportunity at the best price possible.
Brand awareness can be defined as the potential capacity that a consumer has of recognizing or recalling the name of the brand while purchasing a certain category of product. The concept of brand awareness broadly measure the following two dimensions the reminded that fits with the spontaneous recall about a particular brand sans a need of any kind of external stimulus the recall attended that the brand name is knowledge as an offer of a category of products amongst a set of suggested brands Brand awareness is an effective tool that helps to make a predominant selection of product consumers without experience of use of the product and stops experimentation with new products and brands (Hoyer& Brown, 1990). Thus brand awareness acts just as an antecedent to the creation of brand image are in the origin of (Keller, 1993). These two equally influence a consumer to build an assured image in their minds and take a final call in purchase.
Advanced Methods of Marketing Research BRAND IMAGE Kotler defined that A brand is a name, term, design, symbol, or other feature that distinguishes products and services from competitive offerings. As per the Aakers definition, the brand is a specific name or mark, and can be used for distinguishing with competitor's products and servers. A research has shown that a brand image should be based on brand concept-image, which can be built in the following three benefits: Functional: actual benefits from using a product or service, concentrating on satisfying consumers basic needs. Symbolic: added value of a product or services, stressing the ability to fulfill consumers inner needs and self-image Experiential: subjective experience from using the product or service. The brand image is important in marketing because the brand image is considered as the clue of a kind of information PURCHASE INTENTION Purchase intention means probably attempting to buy a product. According to Kotler, consumer behavior occurs when consumers are stimulated by external factors and come to a purchase decision based on their personal characteristics and decision making process. These factors take note of choosing a product, brand, a retailer, timing, and quantity. This means consumers purchasing behavior is triggered by their choice of product and brand. Consumers purchase intentions are always preceded by consumer perceived value and perceived benefit. Hence, the research chooses purchasing intention to be a good indicator of consumers making a buying decision and help to understand whether or not the brand image will significantly influence consumers purchasing intention. Study shown that people comparatively purchase those with which they are familiar and the products with good brand image because the good brand image can make one feel at ease and reliable.
RESEARCH DESIGN
PREPARATORY RESEARCH The research was started with preparatory research. We have explored many resources about footwear industry which included previous marketing research reports on footwear industry and prominent websites from where we were able to get Macroeconomic Information, Demographic Information as well as Company Profiles which has enabled us to build a quite accurate market overview. SECONDARY RESEARCH After the first step we conducted secondary research that ensured us that we were always fully up-todate with the latest industry events and trends, aggregates and analyzes a number of secondary information sources during our research that includes: National/Governmental statistics International data (official international sources) National and International trade associations Chapter: Objectives and detailed methodology
Advanced Methods of Marketing Research Broker and analyst reports Company Annual Reports Business information libraries and databases After the second step, which involved the collection of secondary data, we collected Primary data by directly interviewing various segments of population and conducting a focus group discussion. The type of questions asked while collecting primary data was mostly open ended. After the above 2 steps, we designed a comprehensive Questionnaire which was going to be our primary source of Data collection. Link to questionnaire is https://spreadsheets3.google.com/spreadsheet/viewform?hl=en_US&formkey=dHdpUkJWTGVFa1ZyRm ZyR3BtMzZiVFE6MQ#gid=0
INFORMATION NEEDS
The information needed for the project is related to the essential factors for customers influencing their purchasing behavior. We needed the importance they assign to each of those essential factors while selecting a footwear brand. We also required the perceptions of customers about foreign brands and national brands judging on those factors. DATA COLLECTION FROM SECONDARY RESOURCES A huge amount of secondary resources about the footwear sector were available on the web. We have taken the help of previous marketing research reports on this industry to know the most essential factors influencing the purchasing behavior. This provided us a basic overview of the factors that play a significant role in buying behavior of consumers. DATA COLLECTION FROM PRIMARY RESOURCES We have interviewed various segments of the people in depth to know their preferences regarding the purchase decisions to select a footwear brand. We have also used an online spreadsheet to interview people of various income groups and involved in various occupations. From FGD: Quality, Durability, Availability, Price and Design/fashion came out to be important criterion on which people judge brands whether National or Foreign. Chapter: Information needs Our objective of exploratory research is to find the questions that need to be included in the questionnaire for descriptive research.
10
Q2. Given that the price, quality and other parameters are same for two brands and you know that one Brand is National and other is a foreign brand will you will go for National Brand? Respondents agreed that brand image rules and it doesnt matter whether its a National or Foreign brand. They said Availability is important criteria which influences their buying decision.
Q3. On what parameters do you decide to go for a particular brand? Following answers came out: a. Quality : All agreed b. Durability: All Agreed c. Design/Fashion: All Female Respondents Agreed but Male respondent were Indifferent d. Price: All Agreed. One thing that came out was that a Brand is a package of all the above mentioned traits.
Q4. Does Occasions influence purchase of national or foreign brands? Here the important outcome was that respondents agreed that local brands rather national brands as such are preferred in special occasions. Like in Durga Puja, people prefer to buy clothes from local branded shops for traditional wear.
Q5. Do we repeat ourselves when it comes to purchasing form a particular brand? Respondents agreed that most often than not repeat doesnt happen, people often look for something new if they dont find then they repeat themselves. Few agreed that the brand they like will be given first shot. What is near prevails (Availability) unless until its a very special occasion. High value Items get repeat purchase.
Q6. Do your parents (to get insight into purchasing decision of Old Aged people) show some kind of Brand Loyalty? Mostly respondents agreed that parents were more quality and price conscious and brand awareness itself is lacking in them leave apart the loyalty.
11
Q8. Is there any difference when you purchase for yourself and when you purchase for others (Like Gift)? All respondents agreed that they are very brand conscious when they purchase for others.
FINDINGS
Following were the important findings of the meeting: There is no differentiation as far as National and foreign brands are concerned especially in Apparel segment. Quality, Durability, Availability, Price and Design/fashion came out to be important criterion on which people judge brands whether National or Foreign. People dont associate any patriotism with National Brand. People are more conscious about brands when they purchase for others. Repeat (Brand Loyalty) doesnt happen as people often look for something new if they dont find then they repeat themselves. Sometime people do give the brand they like the first shot. But then what is near prevails (Availability) unless until its a very special occasion. Certain occasion (Occasions with traditional value like Puja) do prompt purchase from local brands but nothing about national brands.
DESCRIPTIVE RESEARCH
We prepared a comprehensive questionnaire using the inputs of preparatory and exploratory research. The soft copies of the surveys were mailed to various sections of the people. We have distributed the hard copies of the questionnaire to the people by physical access like students of NIT, Calicut. The questionnaire was a prepared exhaustively as this is the primary source of data collection for the project. The questions were designed to do a comparative study between the footwear brands available in. The questions helped us to know their perceptions about national and international footwear brands. SCALING TECHNIQUES For our questionnaire, we have asked the users to rank the factors in the order of importance. The users were asked to rank their preference for Quality, Durability, Availability, Price and Design/fashion on national and international footwear brands. This design facilitated us to do the comparative study easily.
12
Advanced Methods of Marketing Research understandable and simple. Ambiguous words in the questions were avoided. We have made sure that the questions were framed in the most polite language possible. PRETESTING The pretesting was done with some selected contacts and friends. Minor modifications were done based on the feedback provided by them. We have eliminated some questions, which were redundant and reduced the size of the questionnaire. We made sure that the brand related information is at the first page of questionnaire and second page consists of demographic information about consumers after the feedback received.
SAMPLING TECHNIQUE
The target population comprises of all the online users that can be approached along with distributing hard copies to individuals from NIT Calicut and IIM Kozhikode. We have selected our samples from individuals residing pan India; accessibility was accomplished by uploading an online questionnaire. We have used a combination of Quota sampling and convenience sampling. The convenience sampling was adopted for ease of administering and analyzing. As majority of Indian population is below the age of 30 years, we ensured that our samples comprise of that section in majority. Certain limitations were also taken into consideration while selecting these two techniques. We have taken care to reduce the bias arising out of these techniques .We have not only included our friends in the samples but made sure that samples contain diversified set of people who are not our friends. We have taken the help of social networking sites to do this.
FIELDWORK
We have mailed the soft copies of questionnaires to most of the contacts. We also made use of certain social networking sites like orkut, face book to get the questionnaires answered. We have met the people of IIM Kozhikode and NIT Calicut to distribute the hard copies of questionnaires filled.
13
Advanced Methods of Marketing Research inconsistent information corresponding to the criteria as essential or missed answering some of the questions) CODING Users were asked to rank the factors in the order of importance. We have converted the rank into the score. The scores are assigned according to the table below RANK 1 2 3 4 5 SCORE ASSIGNED 1 2 3 4 5
We have prepared a codebook, which contains the code we used for each variable and item of data in each question The codebook was used as reference while doing the data analysis.
14
Percent
100.0
15
Advanced Methods of Marketing Research cluster solutions while leaving the final selection of the best cluster solution to none hierarchical procedures.
Agglomeration Schedule Cluster Combined Stage 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Cluster 1 15 23 25 7 30 5 13 15 8 5 18 36 22 24 3 17 2 6 7 9 5 37 11 2 7 5 17 4 5 8 3 1 2 7 5 2 4 4 1 1 Cluster 2 29 26 34 33 31 21 16 19 25 20 41 39 38 32 22 23 35 30 10 18 15 40 27 12 9 14 28 24 6 36 17 3 13 8 11 37 5 7 2 4 Coefficients .000 .000 .500 1.000 1.500 2.000 2.500 3.167 4.000 4.833 5.833 6.833 7.833 8.833 9.833 11.167 12.667 14.167 15.667 17.333 19.333 21.833 24.333 26.833 29.500 32.357 35.274 38.274 42.317 46.383 50.705 55.258 59.958 67.134 75.234 84.963 96.529 113.746 138.392 185.366 Stage Cluster First Appears Cluster 1 0 0 0 0 0 0 0 1 0 6 0 0 0 0 0 0 0 0 4 0 10 0 0 17 19 21 16 0 26 9 15 0 24 25 29 33 28 37 32 39 Cluster 2 0 0 0 0 0 0 0 0 3 0 0 0 0 0 13 2 0 5 0 11 8 0 0 0 20 0 0 14 18 12 27 31 7 30 23 22 35 34 36 38 Next Stage 8 16 9 19 18 10 33 21 30 21 20 30 15 28 31 27 24 29 25 25 26 36 35 33 34 29 31 37 35 34 32 39 36 38 37 39 38 40 40 0
16
Advanced Methods of Marketing Research DEFINING THE NUMBER OF CLUSTERS: ELBOW RULE The agglomeration coefficient is particularly useful in determining the number of clusters. Small coefficients indicate that fairly homogemous clusters are being merged. In contrast when the two different clusters are joined it results in a large coefficient. Each combination of clusters results in increased heterogenity, so we focus on large percentage in coefficient. There is a sudden kick in the coefficients of agglomeration schedule which would help us to determine the number of clusters that can be defined. Given below is the scree diagram plotted through MS Excel which shows that the sudden jump in the values of coefficient is at 39th step. So, the number of clusters would be 41-39 = 2 clusters. DEFINING THE NUMBER OF CLUSTERS: DENDROGRAM
At the last 3 stages of dendrogram, the clusters are being combined at large distances. Therefore it appears that 3 cluster solutions is appropriate.
17
Advanced Methods of Marketing Research ANOVA From the ANOVA table, we can conclude that for F test, there are three variables i.e. X4[Packaging] X5[Price] X8[Availability], Which are not significant, so our assumption of two clusters solution is not the best.
X3 X4 X5 X6 X7 X8 ANOVA Cluster Mean Square 16.919 1.873 1.172 6.669 32.559 .003 df 1 1 1 1 1 1 Error Mean Square .194 .784 .610 .433 .550 .664 df 39 39 39 39 39 39 F 87.182 2.390 1.920 15.397 59.222 .004 Sig. .000 .130 .174 .000 .000 .947
So the numbers of The F tests should be used only for descriptive purposes because the clusters have been clusters in analysis chosen to maximize the differences among cases in different clusters. The observed significance levels are not corrected for this and thus cannot be interpreted as tests of the were increased to hypothesis that the cluster means are equal. 3 and were accepted when the ANOVA was carried out because only the variable X5 was insignificant but rest were significant. Hence below is given the ANOVA table for 3 clusters which clearly depicts that all the variables for F test are significant.
ANOVA Cluster Mean Square X3 X4 X5 X6 X7 X8 7.062 4.493 1.678 4.096 17.887 4.713 df 2 2 2 2 2 2 Error Mean Square .273 .617 .569 .404 .480 .434 df 38 38 38 38 38 38 F 25.897 7.281 2.950 10.127 37.293 10.870 Sig. .000 .002 .064 .000 .000 .000
18
Once the clusters are joined, they are never separated in the clustering process. We have selected Wards method to aggregate the clusters with minimum distances one by one, still the non-hierarchical clustering methods hold the advantage of being able to optimize the clustering solution by reassigning observations until minimum heterogeneity within the clusters is achieved. Thus the primary element of using non-hierarchical technique is to improve the results from hierarchical procedure.
Final Cluster Centers Number of Cases in each Cluster Cluster 1 2 3 Valid Missing 12.000 21.000 8.000 41.000 .000 X3 X4 X5 X6 X7 X8 Iteration History Iteration 1 2 3 4 1 2.306 .137 .180 .000 2 2.616 .172 .115 .000
a
Cluster 1 4 4 4 4 4 4 2 3 3 4 3 2 4 3 4 2 4 4 4 3
a. Convergence achieved due to no or small change in cluster centers. The maximum absolute coordinate change for any center is .000. The current iteration is 4. The minimum distance between initial centers is 5.292.
If we closely check the clusters, we can make the following observations, 1) Cluster 1consists of 12 people who give high emphasis on quality, durability, fashion & availability. They give moderate emphasis on packaging. 2) Cluster 2 consists of 21 people who have high preference for quality & moderate preference for fashion & low preference for packaging 3) Cluster 3 consists of 8 people who have high preference for durability & availability. But they have low preference for packaging & fashion.
19
The table is divided into between the group effects and within the group effects. The between the group effect is overall experimental effect. In this we can see that the SSM (Total sum of squares for model) for Quality is 5.364. The SSM gives the total experimental effect whereas mean square for the model has average experimental effect. The table tells us how unsystematic variation exists. The test whether the group means are same or not are given as F-test ratio which is computed as F-ratio = 5.364/.362 = 14.814 The final value with significance level tells us if this event can occur by chance, but here p .05. This we can say that the effect of quality between three groups is significantly different. We dont know which group mean is significantly different, but that is indicated through Post-Hoc. X6 Durability, X7 Fashion, X8 Availability can be analyzed on similar line. But consider X4 i.e. Packaging where the SSM (Total sum of squares for model) is 2.260. The test whether the group means are same or not are given as F-test ratio which is computed as F-ratio = 2.260/.735 = 3.076
20
Advanced Methods of Marketing Research The final value with significance level tells us if this event can occur by chance, but here p .05. This we can say that the effect of quality between three groups is not significantly different. POST HOC - SCHEFFE
Multiple Comparisons Scheffe Depend (J) ent (I) Ward Ward Mean Difference Variable Method Method (I-J) X3 1 2 3 X4 1 2 3 X6 1 2 3 X7 1 2 3 X8 1 2 3 2 3 1 3 1 2 2 3 1 3 1 2 2 3 1 3 1 2 2 3 1 3 1 2 2 3 1 3 1 2 .625 1.279
*
95% Confidence Interval Std. Error .311 .243 .311 .256 .243 .256 .444 .347 .444 .365 .347 .365 .335 .262 .335 .276 .262 .276 .360 .281 .360 .296 .281 .296 .346 .270 .346 .284 .270 .284 Sig. .147 .000 .147 .050 .000 .050 .061 .230 .061 .424 .230 .424 .004 .001 .004 .866 .001 .866 .216 .000 .216 .000 .000 .000 .000 .245 .000 .003 .245 .003 Lower Bound -.17 .66 -1.42 .00 -1.90 -1.31 -.04 -.28 -2.22 -1.41 -1.49 -.45 .34 .38 -2.05 -.85 -1.72 -.55 -.27 1.48 -1.56 .79 -2.91 -2.30 .62 -.23 -2.38 -1.76 -1.15 .31 Upper Bound 1.42 1.90 .17 1.31 -.66 .00 2.22 1.49 .04 .45 .28 1.41 2.05 1.72 -.34 .55 -.38 .85 1.56 2.91 .27 2.30 -1.48 -.79 2.38 1.15 -.62 .23 1.76 -.31
-1.196 -1.048
-.148
*
-.643
* * * *
-1.549
-.462
*
21
Advanced Methods of Marketing Research Scheffe multiple comparisons show that: For Quality(X3) group means of cluster 1 and cluster 2 are insignificant. For Packaging(X4) the three group means are not significantly from each other For Durability(X6) except cluster 2 and 3 the other groups are significantly different from each other For Fashion (X7) mean of cluster 3 is significantly different from 1 & 2 but means of cluster 1 & 2 are similar
ANOVA Sum of Squares X33 Between Groups Within Groups Total X35 Between Groups Within Groups Total X36 Between Groups Within Groups Total X37 Between Groups Within Groups Total X38 Between Groups Within Groups Total .345 3.265 3.610 .905 11.582 12.488 3.873 26.176 30.049 .014 .962 .976 .384 11.128 11.512 df 2 38 40 2 38 40 2 38 40 2 38 40 2 38 40 .192 .293 .656 .004 .007 .025 .278 .019 1.936 .689 2.811 .043 .453 .305 1.485 .029 Mean Square .172 .086 F 2.006 Sig. .039
From Scheffes Test, Group 1 corresponds to Best Attributes Group 2 corresponds to Moderate Attributes Group 3 corresponds to Low Attributes
22
CLUSTER PROFILING Group 1: 19.5% of the females who mostly are students who fall in the income bracket of 0-2 lakhs & are single reside mostly in city Group 2: 17% of the females who mostly are students but may also be employed who fall in the income bracket of 0-2 lakhs & are single reside mostly in towns. Group 3: 63% of the males who mostly are also employed who fall in the income bracket of greater than 4 lakhs & mostly are single reside in cities.
X14-What attributes do you consider most important while purchasing Foreign Branded Footwear? [Availability]
23
24
DEFINING THE NUMBER OF CLUSTERS: ELBOW RULE Given below is the scree diagram plotted through MS Excel which shows that the sudden jump in the values of coefficient is at 38th step. So, the number of clusters would be 41-38 = 3 clusters.
At the last two stages of dendrogram, the clusters are being combined at large distances. Therefore it appears that 3 cluster solution is appropriate. Chapter: Cluster Analysis
2011 Group I | Indian Institute of Management, Kozhikode | PGP II
25
26
Advanced Methods of Marketing Research K-MEANS CLUSTERING We will now use K-Means Clustering technique is to improve the results from hierarchical procedure.
ANOVA Cluster Mean Square X9 X10 X11 X12 X13 X14 . Final Cluster Centers Cluster 1 X9 X10 X11 X12 X13 X14 2 2 2 2 2 2 2 4 3 2 4 4 3 3 5 4 4 3 4 Valid 5 Missing 4 .000 41.000 15.000 Cluster 1 2 Number of Cases in each Cluster 2.000 24.000 8.666 11.880 16.129 7.507 9.165 4.064 df 2 2 2 2 2 2 Error Mean Square .502 1.005 .757 .495 .389 .669 df 38 38 38 38 38 38 F 17.279 11.820 21.306 15.180 23.545 6.072 Sig. .000 .000 .000 .000 .000 .005
If we closely check the clusters, we can make the following observations (considering post hoc test), 1) Cluster 1consists of 2 people who give low preference for all the attributes. 2) Cluster 2 consists of 24 people who have high emphasis for durability while moderate preference for all other attributes. 3) Cluster 3 consists of 15 people who have moderate preference for availability but high preference for all other attributes. Chapter: Cluster Analysis
27
The table is divided into between the group effects and within the group effects. The between the group effect is overall experimental effect. In this we can see that the SSM (Total sum of squares for model) for Quality is 5.364. The SSM gives the total experimental effect whereas mean square for the model has average experimental effect. The table tells us how unsystematic variation exists. The test whether the group means are same or not are given as F-test ratio which is computed as F-ratio = 5.364/.362 = 14.814 The final value with significance level tells us if this event can occur by chance, but here p .05. This we can say that the effect of quality between three groups is significantly different. We dont know which group mean is significantly different, but that is indicated through Post-Hoc.X6 Durability, X7 Fashion, X8 Availability can be analyzed on similar line. But consider X4 i.e. Packaging where the SSM (Total sum of squares for model) is 2.260. The test whether the group means are same or not are given as F-test ratio which is computed as F-ratio = 2.260/.735 = 3.076 The final value with significance level tells us if this event can occur by chance, but here p .05. This we can say that the effect of quality between three groups is not significantly different.
28
3) 4) 5) 6) 7)
29
From the regression output we can see that regression between these independent and dependent variable is significant (<5%) ESTIMATED MULTIPLE REGRESSION EQUATION E(y) = (-0.297)*X38+ (-0.14)*X33+ (0.009)*X34+ (-0.180)*X36+ (-0.531)*X37+ (-0.495)*X32+ (0.125)*X29+ e Since only two independent variables (X32 and X36) are significant for this dependent variable, hence the correct estimated multiple regression equation would be: E(y) = (-0.180)*X36+ (-0.495)*X32 + e Chapter: Multiple Regression Analysis Other factors are not much significant to consider into regression equation. From coefficients table we can see that special occasions and Income bracket are the most dominant factor in deciding the frequency of purchase of national brand. We can see that there is a high amount of correlation exists between these variables. X36(Income) is -vely correlated with frequency of purchase of national branded footwear. X32(Special events) is vely correlated frequency of purchase of national branded footwear.
30
Points clustered closely around a line show a strong correlation. The line is a good predictor (good fit) with the data. The more spread out the points, the weaker the correlation, and the less good the fit. The line is a REGRESSSION line (Y = bX + a) Hence we can see that there is a high amount of correlation exists between these two variables.
3) 4) 5) 6) 7)
Sex Age Income Bracket Marital Status Living In city/town/village? Chapter: Multiple Regression Analysis
31
From the regression output we can see that regression between these independent and dependent variable is significant (<5%) ESTIMATED MULTIPLE REGRESSION EQUATION E(y) = (-0.455)*X38+ (-0.405)*X33+ (0.013)*X34+ (0.115)*X36+ (-0.524)*X37+ (0.449)*X32+ (0.117)*X29+ e Since only two independent variables (X32 and X36) are significant for this dependent variable, hence the correct estimated multiple regression equation would be: E(y) = (-0.455)*X38+ (0.449)*X32 + e Other factors are not much significant to consider into regression equation. From coefficients table we can see that special occasions and living in city/village or town (place) is the most dominant factor in deciding the frequency of purchase of foreign brand. We can see that there is a high amount of correlation exists between these variables. X38(Living in) is -vely correlated with frequency of purchase of foreign branded footwear. X32(Special events) is +vely correlated frequency of purchase of foreign branded footwear. Chapter: Multiple Regression Analysis
32
We can see that there is a high amount of correlation exists between these two variables. INFERENCE Income motivates consumers to buy foreign brands footwear On special occasions consumers prefer foreign brands instead of national brands footwear Location of consumers also motivates them to go for national or foreign brands o Village/Town: Mostly go for National brands o City: Large population go for Foreign brands
FACTOR ANALYSIS
It is statistical technique that analyzes the relationship among a large number of variables to determine a set of common underlying dimensions Conditions for Factor Analysis
33
As from left-hand table we see The Bertletts test passed, the Chi-square value is low and KMO stats > 0.5 so we say that factor analysis is feasible for the given dataset.
From Total Variance Explained table we find First four factors cumulative squared loading is higher than 60 so we consider four main factors exist.
34
Advanced Methods of Marketing Research From right-hand table we conclude Factor 1 has high coefficients for variables X24,X25,X26,X27 hence we can label this factor as product satisfaction Factor 2 has high coefficients for variables X3,X6,X7 hence we can label this factor as value for money Factor 3 has high coefficients for variables X4,X5,X8 hence we can label this factor as economical Factor4 has high coefficient for variable X19 hence we can label this factor as brand loyalty From above analysis we conclude that consumers purchasing intention for national brand is influenced by value for money, product satisfaction, economical and brand loyalty.
35
As from left-hand table we see The Bertletts test passed, the Chi-square value is low and KMO stats > 0.5 so we say that factor analysis is feasible for the given dataset.
From above right-hand table we find First three factors cumulative squared loading is higher than 60 so we consider four main factors exist. From the left-hand table we conclude Factor 1 has high coefficients for variables X20,X21,X22 hence we can label this factor as product satisfaction Factor 2 has high coefficients for variables X10,X12,X13 hence we can label this factor as value for money Factor 3 has high coefficients for variables X14,X18,X23 hence we can label this factor as brand image From above analysis we conclude that Consumers purchasing intention for foreign brand is influenced by brand image, value for money and product satisfaction Chapter: Structure Equation Modeling
36
From test of validity we find out AVE for PS and VFM > .5 but for Eco <.5 ,so model is not convergent valid AVE for all constraints are greater than inter-construct correlation squared , so model is discriminant valid From test of reliability we find out CF for PS and VFM > .7 but for Eco <.7 , so model is not fully reliable As there was no option of improving the model as all metrices for improvement were showing good results already. Although model was a good fit and discriminant valid it was not convergent valid or reliable. We conclude consumers product satisfaction for national brands is strongly influenced by products value for money but they are not so price sensitive for national brands. Chapter: Structure Equation Modeling
37
THE ANALYSIS From the goodness-of-fit test of the model we find out Chi-square=24, CMIN/DF= 1.001 (<2) hence we can say the model is a good fit GFI= .887(~.9), TLI= .778, CFI= 1.00(>.9)and RMSEA= 0.005 (.08) hence it confirms model is good fit
From test of validity we find out AVE for PS and VFM > .5 but for BI <.5 ,so model is not convergent valid Both inter-construct correlation squared for BI are greater than AVE , so model is was not discriminant valid From test of reliability we find out CF for PS and VFM > .7 but for BI <.7 , so model is not fully reliable As there was no option of improving the model as all metrices for improvement were showing good results already. Although model was a good fit it was not convergent valid nor discriminant valid nor reliable. We conclude consumers product satisfaction for foreign brands is strongly influenced by products value for money but brand image does not influence it.
38
REFERENCES
TEXTBOOK REFERENCES
Marketing Research 6e, An Applied Orientation Naresh K Malhotra, Satyabhushan Dash Multivariate Data Analysis 6e Hair, Black, Babin, Anderson, Tatham
WEB REFERENCES
A journal by Angel F. Villarej, Francisco J. Rondn, Manuel J. Snchez http://www.slideshare.net/BiratSharma/cluster-spss-week7 http://core.ecu.edu/psyc/wuenschk/MV/FA/FA.doc www.wikipedia.com Academic-papers.org/ocs2/session/Papers/G2/146.doc by Kuang-Wen Wu and KunChang Wu
DATABASE REFERENCES
39
Chapter: References