Product Innovativeness Trial Ability New Product Marketing Essay

Published Date: 23 Mar 2015

When a new product was launched into an established product class, consumers became aware of the product through the marketing efforts of the firm launching the product and through word-of-mouth communication by current buyers. These two sources of indirect product experience stimulated trial. The decision to repurchase the new product or switch to a competing product was likely to be based on direct product experience. Even if the trial experience was negative and the consumer went back to the competing product, a switch to the new product can still occur because of continued marketing efforts and positive word-of-mouth. Similarly, the consumer who repurchased the new product may switch to the competing product because of continued competitive marketing effort. A new product model should capture this dynamic element of consumer behavior and the effect of marketing variables and word-of-mouth.

Seethaletchumy, Uchenna, Khong, Robert and Kim (2010) have discussed trial ability in their research on Islamic banking services that customers were unable to try it. Therefore to gain the confidence of apprehensive customers it was necessary that they should use this service on trial basis. Rogers (2003) stated that positive relationship was assumed with trial ability and adoption rate and easy trial of an innovation would get better rate of adoption. Seethaletchumy et al (2010) stated that it was crucial to test the trial ability of Islamic banking in Malaysian banking market.

The researchers Banerjee, Wei and Ma (2010) have discussed trial ability in their research abstract as:

Trial ability has been conceptualized in prior research as a belief signifying opportunity to experiment with a technical innovation would facilitate its adoption. It has been found to be a weak predictor and though indications exist of possible significant impact in situations of high perceived risk, it has not received serious academic attention. In this research we argued that in situations of high perceived risk, the belief-based concept of Trial ability without actual evaluation of experimental outcomes was questionable. Based on the Theory of Trying, Expectancy Disconfirmation theory, and prior research on risk and trust in e-business, we developed and validated propositions in the context of B2B e-market transactions by two small firms. Findings indicated that due to high perceived risk of B2B e-market transactions, in contrast to the received notion of Trial ability as a belief-based factor, it was in the nature of an active post-intent recursive process of experimentation involving 'Trying' for trial transactions with controlled risk, execution of 'Trial' transactions and 'Assessment of Trial Outcomes'. Also, in contrast to weak impacts observed in prior research, the trial ability process was found to be a necessary condition for translation of initial intent to adoption.

The researchers Hafizah and Kamil (2009) discussed trial ability with the research model shown in figure 2.1 examining the factors that could influence the E-learning adoption by using Perceived Attributes Theory. Data was collected by taking University Utara Malaysia (UUM) as a sample to empirically test the hypothesized relationships of relative advantages, trial ability and academic specialization. This was concluded that the research model in figure 2.1 showed a logically good fit among the data and empirical outcome established that only relative advantages, trial ability and academic specialization optimistically influenced the adoption decision. Thus, the conclusions had provided evidence of the importance of this in understanding the decision of adoption prior to introducing new online technology and instructional delivery in education.

Figure 2.1 A model for E-learning adoption

(Source: Hafizah Mohamad Hsbollah and Kamil Md. Idris, 2009)

The researcher Tan and Eze (2008) in their study using perceived characteristics of innovations proposed by Roger concluded and stated that:

Findings from Pearson correlation and multivariate regression analysis indicated that following Roger's adoption excluding trial ability and ICT security and confidentiality were significant contributors to ICT adoption intention in Malaysian SMEs. Trial ability and ICT costs were not significantly related to ICT adoption. One possible reason that trial ability was not a significant factor was trial ability of software was not available in Malaysia.

Complexity

Trial ability

Relative Advantage

Observe ability

New Product Adoption

Compatibility

Figure 2.2 Rogers New Product Adoption Model (Persuasion)

After reviewing the literature as above which has provided information about the product innovativeness in combination of different perspective (newness to company, newness to market and newness to customer) and in combination of different factors with and without addition in Rogers's proposed new product adoption model for specific product, business and industry. The proposed research model was presented in figure 2.3 for this study. Identification of the most common variables/dimensions out of product innovativeness and trial ability dimensions which were leading to increase the new product adoption was the purpose of this study. Further relationship between consumer characteristics, product innovativeness and trial ability in new product adoption was tried to find with the help of following proposed research model.

New Product Adoption

Trial ability

Product Innovativeness

Demographic characteristics of the respondents

(Gender, age, income

Education, occupation, marital status.

Figure 2.3 Proposed Research Model

CHAPTER 3: RESEARCH METHODS

In this chapter, method of the study was discussed. This chapter contained method of data collection, sample design, questionnaire design, content validity, reliability of the research instrument and statistical technique used in this research.

3.1 Method of Data Collection

An instrument was developed initially to test the content and evaluation validity for this research. The questionnaire was filled by the five selected respondents. The responses of these respondents were excluded from actual study process. A survey was conducted to determine the effect of product innovativeness and trial ability on new product adoption. The students of various universities in Karachi were a target population. The questionnaires were delivered to students of various universities in Karachi in the presence of researcher to ensure interaction and clarification (if any). Completed questionnaires were collected from the respondents' and data was entered in computer software (SPSS) for quantitative analysis. The computed means were compared to the scale below for interpretation:

Range

Interpretation

0.00 - 1.49

Strongly Disagree

1.50 - 2.49

Disagree

2.50 - 3.49 

Neither agree nor disagree

3.50 - 4.49

Agree

4.50 - 5.00

Strongly Agree

3.2 Sample Design

In order to find out whether there was any effect of product innovativeness and trial ability on new product adoption, a sample size of 108 respondents was selected conveniently from the students of universities in Karachi. To achieve relevant information, certain additional criteria were imposed. The participants' inclusion in sample was that they must have purchased at least a new product (in their point of view) in at the most 30 days earlier. This ensured that the participants can recall their memories to response the questionnaire (instrument). Out of 108 respondents 88 were male and 20 were female which shows 81.5% and 18.5% respectively. Marital Status of the respondent as: single, married and divorced in numbers are 91, 15 and 2 in both genders with percentage as: 84.3%, 13.9% and 1.8% respectively. The Occupational status of the respondent as: student, employed and self employed in numbers are 49, 51 and 8 in both genders with percentage as: 45.4%, 47.2% and 7.4% respectively. The Household average income per month in Pak Rupees of the respondents as: Under Rs. 20,000, Rs. 20,001-40,000, Rs. 40,001-60,000, Rs. 60,001-80,000, Rs 80,001-100,000 and Above Rs. 100,000 in numbers are 21, 37,14,15,7 and 14 in both genders with percentage as: 19.4%, 34.3% ,13.0%.13.9%, 6.5% and 12.9% respectively. The Educational level of the respondent as: Bachelor or equivalent, Master or equivalent and Professional in numbers are 38, 56 and 14 in both genders with percentage as: 35.2%, 51.8% and 13.0% respectively.

3.3 Questionnaire Design

A questionnaire was used as an instrument in this research and was developed after review various studies on product innovativeness, trial ability and new product adoption. This has four sections; new product purchase was first section, product innovativeness was second section, trial ability a third section and demographics (gender, marital status, age, occupation, household average income, education) was fourth section. Likert format was used in structuring question and five choices were provided in each statement from the degree of agreement to disagreement which enabled the respondents to answer easily.

5

Strongly Agree

4

Agree

3

Neither agree nor disagree

2

Disagree

1

Strongly Disagree

3.4 Content Validity of Research Instrument

The instrument was developed after literature review in the area of product innovativeness, trial ability and new product adoption. Different variables of these factors were developed. After thorough discussion with the supervisor two variables for new product adoption ,seventeen variables for product innovativeness, nine variables for trial ability and six information regarding respondent's socio-demographic characteristics were included in the instrument. For content validity of instrument a pilot survey from five respondents was conducted with the request of any suggestion/correction or any ambiguity (if any) in any question. On the basis of assessment and suggestion from the pilot sample respondents the researcher modified the content to ensure the survey comprehensive for the study sample.

3.5 Reliability Analysis of the Research Instrument (Questionnaire)

Cooper (2003) stated that Reliability means many things to many people, but in most contexts the notion of consistency emerges. Reliable instruments are robust; they work well at different times under different conditions. Various tests to measure the instrument's reliability are available in statistics but Cronbach's Coefficient Alpha test of reliability is applied in this study. The results were as under:

Table: 3.1

Reliability Statistics

Cronbach's Alpha

N of Items

0.662

26

0.611

17

0.632

9

The Table: 3.1 showed the reliability statistics of the instrument for 17 items for product innovativeness and 9 items for trial ability. For total 26 items Cronbach's Alpha was 0.662 and separately for 17 items of product innovativeness was 0.611 and for 9 items of trial ability was 0.632 which were satisfactory.

3.6 Statistical Technique

Two statistical techniques were used in this research to answer the research questions. To find the answer of first question analysis of variance (ANOVA) technique was used. This technique was used because it helped in identification of dimensions/ variables of product innovativeness and trial ability which were leading in adoption of innovations on the basis of comparing their means. The second question was answered by applying Logistic Regression (binary/multinomial). Logistic Regression was used to find out the relationship/ behavior of the product innovativeness / trial ability dimensions across the different demographic groups. The basis to use Logistic Regression was explained by Hair, Black, Babin, Anderson and Tatham (2006) in these words, Logistic Regression was a special form of regression that was formulated to predict and explain a binary (two group) categorical variable. Logistic Regression variate was similar to the variate in multiple regressions. The variate represented a single multivariate relationship with similar to regression coefficients that indicated the relative impact of each predictor variable.

CHAPTER 4: DATA ANALYSIS & RESULTS

4.1 Respondent's Profiles

A survey was conducted by the researcher on a developed instrument comprising two basic question about purchase and period of purchase to check the latest product purchased. Then seventeen dimensions were mentioned about the product innovativeness characteristics and nine dimensions were mentioned about the trial-ability characteristics. The information about six demographic (gender, marital status, age group, occupational status, household average income, educational level) of the respondents were collected. Because convenience sampling technique was adopted so 100% response rate achieved from 108 respondents. Respondent's profiles results were presented in table 4.1 and visually can be seen from graph 4.1- 4.6 respectively.

Table 4.1

Respondents Profiles

Gender of the Respondent

Female

Male

Total

%age

Marital Status of the respondent

Single

16

75

91

84.3

Married

3

12

15

13.9

Divorced

1

1

2

1.8

Occupational status of the respondent

Student

14

35

49

45.4

Employed

4

47

51

47.2

Self Employed

2

6

8

7.4

Household average income per month In Pak Rupees of the respondent

Under Rs. 20,000

0

21

21

19.4

Rs. 20,001-40,000

5

32

37

34.3

Rs. 40,001-60,000

2

12

14

13.0

Rs. 60,001-80,000

7

8

15

13.9

Rs 80,001-100,000

3

4

7

6.5

Above Rs. 100,000

3

11

14

12.9

Educational level of the respondent

Bachelor or equivalent

9

29

38

35.2

Master or equivalent

9

47

56

51.8

Professional

2

12

14

13.0

Graph 4.1

Gender of the respondents

Graph 4.2

Gender of the respondents with marital status

Graph 4.3

Gender of the respondents with age group

Graph 4.4

Gender of the respondents with occupational status

Graph 4.5

Gender of the respondents with household income

Graph 4.6

Gender of the respondents with educational level

4.2 Descriptive Statistics

Table 4.2

Descriptive Statistics of Product Innovative Dimensions

N

Mean

Standard.

Deviation

Standard. Error

95% Confidence Interval for Mean

Minimum

Maximum

Lower Bound

Upper Bound

1

108

3.99

1.063

.102

3.79

4.19

1

5

2

108

3.62

1.030

.099

3.42

3.82

1

5

3

108

2.48

1.249

.120

2.24

2.72

1

5

4

108

3.02

1.238

.119

2.78

3.25

1

5

5

110

3.66

.979

.093

3.48

3.85

1

5

6

106

4.12

.739

.072

3.98

4.27

1

5

7

108

3.94

.965

.093

3.76

4.13

1

5

8

108

3.65

.989

.095

3.46

3.84

1

5

9

108

3.67

1.192

.115

3.44

3.89

1

5

10

108

2.44

1.362

.131

2.18

2.70

1

5

11

108

2.64

1.203

.116

2.41

2.87

1

5

12

108

3.77

.943

.091

3.59

3.95

1

5

13

108

3.06

1.277

.123

2.82

3.31

1

5

14

108

2.99

1.204

.116

2.76

3.22

1

5

15

108

3.31

1.336

.129

3.05

3.56

1

5

16

108

3.18

1.393

.134

2.91

3.44

1

5

17

108

3.42

1.375

.132

3.15

3.68

1

5

Total

1836

3.35

1.264

.029

3.29

3.41

1

5

Table 4.3

Descriptive Statistics of Trial ability Dimensions

N

Mean

Standard Deviation

Standard Error

95% Confidence

Interval for Mean

Mini

mum

Maxi

mum

Lower Bound

Upper Bound

1

108

2.81

1.145

.110

2.60

3.03

1

5

2

108

3.37

1.141

.110

3.15

3.59

1

5

3

108

3.16

1.201

.116

2.93

3.39

1

5

4

108

3.74

.980

.094

3.55

3.93

1

5

5

108

2.83

1.308

.126

2.58

3.08

1

5

6

108

3.47

1.256

.121

3.23

3.71

1

5

7

108

3.68

1.126

.108

3.46

3.89

1

5

8

108

3.44

1.154

.111

3.22

3.66

1

5

9

108

3.58

.939

.090

3.40

3.76

1

5

Total

972

3.34

1.184

.038

3.27

3.42

1

5

Descriptive statistics of the instrument for 17 items for product innovativeness with mean, standard deviation and standard error showed in table 4.2 which provided the guidelines for further investigation/ testing of equality of means and variance of these.

Table 4.3 showed the descriptive statistics of the instrument for 9 items for trial ability with mean, standard deviation and standard error which provided the guidelines for further investigation/ testing of equality of means and variance.

4.3 Findings & Interpretation of the Results

In this research, fourteen hypotheses were made on the basis of literature review. To test these hypotheses ANOVA technique was applied to find the answer of first research question with hypothesis 1-2 and Logistic Regression Models (binary /multinomial) were developed and applied to find the answer of second research question with hypothesis 3-14.

4.3.1 Hypothesis (H1)

To test H1 hypothesis that all product innovativeness dimensions for new product adoption have equal means across every group ANOVA technique was applied and the results were as under:

Table 4.4

Test of Homogeneity of Variances

Levene Statistic

Degree of freedom1

Degree of freedom2

Level of Significance.

10.672

16

1819

.000

Table 4.4 showed the significance value less than .05 for the Levene statistics which rejected the hypothesis of homogeneity of variances (equal variances) and accepted that the variances were not equal.

Table 4.5

ANOVA

Sum of Squares

Degree of freedom

Mean Square

F. Statistics

Level of Significance

Between Groups

468.86

16

29.304

21.645

.000

Within Groups

2462.64

1819

1.354

Total

2931.51

1835

Table 4.5 showed the significance value less than .05 for the between groups F statistics which rejected the hypothesis that all product innovativeness dimensions have equal means across every group and accepted that all product innovativeness dimensions have not equal means across every group. It showed that all dimensions of product innovativeness were not equally important to every customer irrespective of nature of product. The mean value of each dimension ≥3.5 was taken as thresh hold and found mean value of the product innovative dimensions number 1,2,5,6,7,8,9 and 12 greater than 3.5 and were important dimensions for new product adoption irrespective of nature of product. (1. This new product was available in a modified version of an existing product; 2. This new product was available in a new version of the existing product; 5. This new product has unique, innovative features; 6.This new product has more satisfying features; 7.This new product was superior in technology / quality; 8. This new product was available in attractive packaging to be purchased; 9.This new product was available by the innovative brand which was not purchased earlier by me; 12. This new product was environment -friendly). These can be verified from Graph 4.7

Graph 4.7

Error bar of mean of Innovativeness dimensions

4.3.2 Hypothesis (H2)

To test H2 hypothesis that all trial ability dimensions for new product adoption have equal means across every group ANOVA was technique was applied and the results were as under:

Table 4.6

Test of Homogeneity of Variances

Levene Statistic

Degree of freedom1

Degree of freedom2

Level of Significance

4.351

8

963

.000

Table 4.6 showed the significance value less than .05 for the Levene statistics which rejected the hypothesis of homogeneity of variances (equal variances) and accepted that the variances were not equal. Table 4.7 showed the significance value less than .05 for the between groups F statistics which rejected the hypothesis that all trial ability

dimensions have equal means across every group and accepted that all trial ability dimensions have not equal means across every group. It showed that all dimensions

Table 4.7

ANOVA

Sum of

Squares

Degree of

freedom

Mean

Square

F.

Statistics

Level of

Significance

Between Groups

100.00

8

12.500

9.547

.000

Within Groups

1260.91

963

1.309

Total

1360.91

971

of trial ability were not equally important to every customer irrespective of nature of product. The mean value of each dimension ≥3.5 was taken as thresh hold and found mean value of the trial ability dimensions number 4,7and 9 greater than 3.5 and were important dimensions for new product adoption irrespective of nature of product. (4. The trial of this new product does not require special equipment; 7. Before deciding on whether or not to adopt a new product, you would need to use it on a trail basis; 9. You should be permitted to use a new product on a trial basis long enough to see what it can do). These can be verified from Graph 4.8

Graph 4.8

Error bar of mean of Trial ability dimensions

4.3.3 Hypothesis (H3)

To test H3 hypothesis (The behavior of product innovativeness dimensions is the same across the gender groups.) Logistic Regression (binary) technique was applied and the results were as under:

Table 4.8

Model Summary

Step

-2 Log likelihood

Cox and Snell R Square

Nagelkerke R Square

1

84.514(a)

.453

.604

Table 4.8 showed the logistic regression model summary for different R square methods. Table 4.9 showed the final results of logistic regression model which included product innovativeness dimension 7,8,14 with interaction of occupational status of the respondents as significant in new product adoption having odd or Exponent (B) values 2.136, 0.433, 1.631, 0.727, 3.962.These dimensions were; 7. This new product was superior in technology / quality; 8. This new product was available in attractive packaging to be purchased; 14. On pack offer was source of attraction to buy this new product; and interaction with occupational status of the respondents was also significant. As the values of Exp (B) of variable 7, 14 and interaction with occupation 2

Table 4.9

Variables in the Equation

Beta coefficient

Standard Error

Wald Statistics

Degree of freedom

Significance level

Exp(B)

Step 1(a)

Prodinnov7

.759

.283

7.202

1

.007

2.136

Prodinnov8

-.838

.353

5.634

1

.018

.433

Prodinnov14

.489

.247

3.930

1

.047

1.631

Occupation

6.694

2

.035

Occupation(1)

-.319

.894

.127

1

.721

.727

Occupation(2)

1.377

1.015

1.840

1

.175

3.962

a. Variable(s) entered on step 1: Prodinnov7, Prodinnov8, Prodinnov14, and Occupation.

(Self employed) were greater than1 which indicated a positive relationship. The values of Exp (B) of variable 8 and interaction with occupation 1 (employed) were less than1 which indicated a negative relationship between both independent variable and predicted probability. Magnitude and probability can be calculated by the following formulae:

Percentage change in odds = (Exp (B) - 1) x 100

Probability = odd/1+odd

The graph 4.9 for the Error Bar of Gender and Product innovativeness 7(This new product was superior in technology / quality) was presented below. Probability of male respondents was .914 for this independent variable as compared to the female respondents.

Graph 4.9

Error Bar of Gender &Product innovativeness 7

This was interpreted as males were more new product adoptive or inclined to purchase if it was superior in quality or technology than female. The graph 4.10 for the Error Bar of

Gender and Product innovativeness 8 (This new product was available in attractive packaging to be purchased) was presented. Probability of female respondents was 1.00 for this independent variable as compared to the male respondents. This was interpreted as females were more new product adoptive or inclined to purchase if it was available in attractive packaging than male.

Graph 4.10

Error Bar of Gender &Product innovativeness 8

The graph 4.11 for the Error Bar of Gender and Product innovativeness 14 (On pack offer was source of attraction to buy this new product) was presented. Probability of male respondents was .923 for this independent variable as compared to the female respondents. This was interpreted as males were more new product adoptive or inclined to purchase if it has on pack offer as source of attraction to buy this new product than female.

Graph 4.11

Error Bar of Gender &Product innovativeness 14

4.3.4 Hypothesis (H4)

To test H4 hypothesis (The behavior of trial ability dimensions is the same across the gender groups.) Logistic Regression (binary) technique was applied and the results were as under:

Table 4.10

Model Summary

Step

-2 Log likelihood

Cox and Snell R Square

Nagelkerke R Square

1

89.205(a)

.124

.201

Table 4.11

Variables in the Equation

Beta

coefficient

Standard

Error

Wald Statistics

Degree of freedom

Level of Significance

Exp(B)

Step 1(a)

Trialability2

-1.055

.332

10.077

1

.002

.348

Trialability3

.574

.248

5.358

1

.021

1.776

Constant

3.538

1.202

8.668

1

.003

34.393

a. Variable(s) entered on step 1: Trialability2, Trialability3.

Table 4.10 showed the logistic regression model summary for different R square methods. Table 4.11 showed the final results of logistic regression model which included trial ability 2, 3, and constant as significant in new product adoption having odd or Exp (B) values 0.348, 1.776, 34.393. These dimensions were; 2. This new product trial involves low risk; 3. The trial of this new product was inexpensive; with constant as significant in the model. The value of Exp (B) of variable 3 was greater than1 which indicated a positive relationship between independent variable and predicted probability. The value of Exp (B) of variable 2 was less than1 which indicated a negative relationship between independent variable and predicted probability. Magnitude and probability can be calculated by the formulae mentioned above.

Graph 4.12

Error Bar of Gender &Trial ability 2

The graph 4.12 for the Error Bar of Gender and Trial ability 2 (This new product trial involves low risk). Probability of female respondents was 0.96583 for this independent variable as compared to the male respondents. This was interpreted as females were more new product adoptive or inclined to purchase if it was available in trial involves low risk than male. In Pakistani culture females were not risk taking as compared to the male.

The graph 4.13 for the Error Bar of Gender and Trial ability 3 (The trial of this new product was inexpensive) was presented. Probability of male respondents was 0.85787 (agree) and 0.82312 (strongly agree) for this independent variable as compared to the female respondents. This was interpreted as males were more new product adoptive or inclined to purchase if it was available in trial involves inexpensive risk than female. In Pakistani culture males were more risk taking as compared to the female. If the trial was inexpensive then males were more inclined to adopt new product.

Graph 4.13

Error Bar of Gender &Trial ability 3

4.3.5 Hypothesis (H5)

To test H5 hypothesis that the behavior of product innovativeness dimensions is the same across the groups of occupation, Logistic Regression (multinomial) was applied.

The results and decision criteria were as under:

Decision criteria:

If the significance level of the Wald statistic is small (less than 0.05) then the parameter is different from 0.

Parameters with significant negative coefficients decrease the likelihood of that response category with respect to the reference category.

Parameters with significant positive coefficients increase the likelihood of that response category with respect to the reference category.

A standard error >2.0 indicates the problems, for example multicollinearity between independent variables

Table 4.12

Pseudo R-Square

Cox and Snell

.747

Nagelkerke

.841

McFadden

.626

The table 4.12 showed that multinomial regression model Pseudo R-Square which was alternate to the R-Square. In this case Cox and Snell value was .747 which showed that dependent variable (Respondent's Occupation) was explained 74.7% by the independent variables/predictors (Product innovative variable 15,16 and demographic characteristics marital status, income, education of the respondents)

Table 4.13

Likelihood Ratio Tests

Effect

Model Fitting Criteria

Likelihood Ratio Tests

-2 Log Likelihood of Reduced Model

Chi-Square value

Degree of freedom

level of

Significance

Prodinnov15

92.559

14.488

2

.001

Prodinnov16

108.760

30.689

2

.000

Mari status

123.204

45.133

4

.000

Income

118.349

40.279

10

.000

Education

97.116

19.046

4

.001

The Table 4.13 showed statistical significance of the relationship between respondent occupation and predictors (Product innovative variable 15, 16 and demographic characteristics marital status, income, education of the respondents).

Occupational status of the respondent(a)

Beta coefficient

Standard

Error

Wald

Degree of freedom

Significance level

Exp(Beta coefficient)

Student

Prodinnov15

-19.907

376.364

.003

1

.958

2.26E-009

Prodinnov16

49.078

888.924

.003

1

.956

2E+021

[Maristatus=1]

58.136

4910.177

.000

1

.991

2E+025

[Maristatus=2]

-116.604

6323.545

.000

1

.985

2.29E-051

[Maristatus=3]

-257.872

25489.16

.000

1

.992

1.02E-112

[Income=1]

-112.857

4799.463

.001

1

.981

9.70E-050

[Income=2]

56.732

4324.541

.000

1

.990

2E+024

[Income=3]

137.597

1.167

13903.84

1

.000

6E+059

[Income=4]

58.415

5015.769

.000

1

.991

2E+025

[Income=5]

34.639

6957.046

.000

1

.996

1E+015

[Income=6]

0(c)

.

.

0

.

.

[Education=2]

56.350

2770.952

.000

1

.984

3E+024

[Education=3]

-50.375

2573.857

.000

1

.984

1.33E-022

[Education=4]

0(c)

.

.

0

.

.

Employed

Prodinnov15

-20.541

376.364

.003

1

.956

1.20E-009

Prodinnov16

49.632

888.924

.003

1

.955

4E+021

[Maristatus=1]

58.374

4910.177

.000

1

.991

2E+025

[Maristatus=2]

-97.372

5070.831

.000

1

.985

5.15E-043

[Maristatus=3]

-243.647

24298.429

.000

1

.992

1.53E-106

[Income=1]

-109.521

4799.463

.001

1

.982

2.73E-048

[Income=2]

58.906

4324.541

.000

1

.989

4E+025

[Income=3]

139.031

.000

.

1

.

2E+060

[Income=4]

59.231

5015.769

.000

1

.991

5E+025

[Income=5]

35.689

6957.046

.000

1

.996

3E+015

[Income=6]

0(c)

.

.

0

.

.

[Education=2]

54.249

2770.952

.000

1

.984

4E+023

[Education=3]

-52.591

2573.857

.000

1

.984

1.45E-023

[Education=4]

0(c)

.

.

0

.

.Table 4.14

Parameter Estimates

Thus hypothesis that the behavior of product innovativeness dimensions is the same across the groups of occupation was rejected. Thus a relationship between respondent's occupations and predictors (Product innovative variable 15, 16 and demographic characteristics marital status, income, education of the respondents) was supported.

From the table 4.14 of parameter estimates multicollinearity was detected from the values of standard errors. A standard error >2.0 indicated the problems, for example multicollinearity between independent variables and could not be interpreted. Standard Error score for respondent's in the household average income group from Rs.40, 001 to 60,000 (Income=3) was less than 2 for Student relative to Self Employed and Employed relative to Self Employed and showed significance less than .05 can be interpreted as positive relationship that if it was increased this will expected to result in increased in the multinomial log-odds of Student relative to Self Employed and Employed relative to Self Employed while other variables were constant in the model. The relationship of these variables was further elaborated from the following graphs:

The graph 4.14 showed that the probability of students with marital status of single relative to self employed was .7749 or 77.49% to purchase new innovative product which offer a free sample ( free sample is important in considering buying new product). For other two categories employed and self employed this variable was not significant.

The graph 4.15 showed that the probability of students with marital status of single relative to self employed was .6952 or 69.52% to strongly disagree to purchase new innovative product because of friend's advice to purchase (friends advice is not important

in considering buying new product for student with marital status of single but this is important for self employed with marital status of single).

Graph 4.14

Error Bar of Occupation (student) & Product Innovativeness 15

With Marital Status of the respondent

Graph 4.15

Error Bar of Occupation (student) & Product Innovativeness 16

With Marital Status of the respondent

The graph 4.16 showed that the probability of employed with marital status of married and divorced relative to self employed was 1.00 or 100% to agree and strongly agree respectively to purchase new innovative product because of friend's advice to purchase (friends advice is important in considering buying new product for employed with marital status of married and divorced).

Graph 4.16

Error Bar of Occupation (employed) & Product Innovativeness 16

With Marital Status of the respondent

The graph 4.17 showed that the probability of self employed with marital status of divorced was 1.00 or 100% to agree to purchase new innovative product because of friend's advice to purchase (friends advice is important in considering buying new product).

The graph 4.18 showed that the probability of students with household average income Rs.40, 000and above relative to self employed are above .90 or 90% strongly agree

Graph 4.17

Error Bar of Occupation (self employed) & Product Innovativeness 16

With Marital Status of the respondent

Graph 4.18

Error Bar of Occupation (student) & Product Innovativeness 15

With Household Average Income of the respondent

Graph 4.19

Error Bar of Occupation (employed) & Product Innovativeness 15

With Household Average Income of the respondent

to purchase new innovative product due to free sample (free sample is important in considering buying new product for students with income group of Rs.40, 000and above).

The graph 4.19 showed that the probability of employed with household average income under Rs.20, 000 relative to self employed were above .962 or 96.2% towards strongly disagree to purchase new innovative product due to free sample (free sample is not important in considering buying new product for employed with income group of under Rs.20, 000 but this is important for self employed of this income group).

The graph 4.20 showed that the probability of students with household average income Rs.100, 000and above relative to self employed were .9579 or 95.79% to disagree to purchase new innovative product due to friend's advice (friend's advice is not important in considering buying new product for students but it is important for self employed).

Graph 4.20

Error Bar of Occupation (student) & Product Innovativeness 16

With Household Average Income of the respondent

Graph 4.21

Error Bar of Occupation (employed) & Product Innovativeness 16

With Household Average Income of the respondent

The graph 4.21 showed that the probability of employed with household average

Income Rs. 80,001 to 100,000 relative to self employed was 1.00 or 100% to neutral to purchase new innovative product due to friend's advice to purchase (friends advice is irrelevant in considering buying new product for employed).

The graph 4.22 showed that the probability of self employed with household average income under Rs. 20,000 was .60 or 60% to disagree to purchase new innovative product due to friend's advice to purchase (friends advice is not important in considering buying new product). Other income groups were below to this value.

Graph 4.22

Error Bar of Occupation (self employed) & Product Innovativeness 16

With Household Average Income of the respondent

The graph 4.23 showed that the probability of students with education level of Master or equivalent relative to self employed was .8211 or 82.11% to purchase new innovative product which offer a free sample ( free sample is important in considering buying new product). For other two categories employed with education level of Bachelor or equivalent was towards disagree and self employed was neutral this variable was not significant.

Graph 4.23

Error Bar of Occupation (student) & Product Innovativeness 15

With Educational Level of the respondent

Graph 4.24

Error Bar of Occupation (student) & Product Innovativeness 16

With Educational Level of the respondent

The graph 4.24 showed that the probability of students with education level of Bachelor or equivalent relative to self employed was .6988 or 69.88% to strongly disagree to purchase new innovative product due to friend's advice ( friend's advice is not important in considering buying new product).

Graph 4.25

Error Bar of Occupation (employed) & Product Innovativeness 16

With Educational Level of the respondent

The graph 4.25 showed that the probability of employed with Professional education level relative to self employed was 1.00 or 100% to strongly agree to purchase new innovative product due to friend's advice ( friend's advice is important in considering buying new product). For self employed it was not significant.

4.3.6 Hypothesis (H6)

To test H6 hypothesis that the behavior of trial ability dimensions is the same across the groups of occupation, Logistic Regression (multinomial) technique was applied and the results were as under:

Table 4.15

Pseudo R-Square

Cox and Snell

.060

Nagelkerke

.072

McFadden

.034

The Table 4.15 showed that multinomial regression model Pseudo R-Square which was alternate to the R-Square. In this case Cox and Snell value wais .060 which showed that dependent variable (Respondent's Occupation) was explained 6% by the independent variables/predictors (Trial ability variable 4: The trial of this new product does not require special equipment). This model was not powerful because only 6% variation was explained by the predictor (Trial ability variable 4 i.e. the trial of this new product does not require special equipment).

The table 4.16 showed statistical significance of the relationship between respondent occupation and predictors (Trial ability variable 4: The trial of this new product does not require special equipment). Thus the hypothesis that the behavior of trial ability dimensions is the same across the groups of occupation was rejected. Thus a relationship between respondent's occupation and predictors (Trial ability variable 4) was supported.

Table 4.16

Likelihood Ratio Tests

Effect

Model Fitting Criteria

Likelihood Ratio Tests

-2 Log Likelihood of Reduced Model

Chi-

Square statistic

Degree of freedom

level of Significance

Intercept

37.995

14.017

2

.001

Trialability4

30.657

6.679

2

.035

From the table 4.17 it can be seen that comparing survey of respondents who were student in occupation relative to those who were self employed in occupation, the probability of the Wald statistic (4.476) for the variable Trial ability 4 was 0.034. Thus the hypothesis that the B coefficient value for Trial ability 4 was equal to zero for this