NZ & GE
Regression: 2005 Model with 2009 Dataset
Model Summary
Model R R Square Adjusted R
Square Std. Error of the Estimate
Intention .921a .848 .846 .464
a. Predictors: (Constant), Trust, Social Norms, Intrinsic, Extrinsic, Product-Specific Attitude
Coefficients
Model Unstandard-ized Coeffi-
cients
Standardized Coefficients t Sig.
B Std. Error Beta
Intention (Constant) .185 .099 1.876 .061 Intrinsic .019 .038 .018 .498 .619 Extrinsic .101 .044 .090 2.280 .023
Social Norms .032 .040 .025 .791 .430
Product-
specific Atti-tude
.911 .042 .861 21.879 .000
Trust -.075 .038 -.063 -1.946 .052
a. dependent variable: Intention
0100200300400500600700
Totally Support Support in some circumstance
Totally oppose
No
. of
resp
on
ses
Response Frequencies of Different GE Applications
13. GE in Food
14. GE in Medicine
Changes in Attitude
The final question concerns a time-series-like analysis of the data across the four years. Answers to this question were sought by using Fishers least
significant difference (LSD) post-hoc ANOVA test in conjunction with line plots of means across time. This allowed for the comparison of means be-
tween each of the four years. General trends were alluded to through the observation of changes in means.
Explanatory
The final stage of analysis involved the discussion for the results observed from which conclusions will be drawn. This will require drawing upon lit-
erature and works in this field. The product of this stage of analysis will be in the form of validation of Smalls model or the creation of a new model
and a set of recommendations for future research or strategies to be implemented.
Empirically
determined
standardised
regression
coefficient
weighted
beliefs or
belief factors
Perceived prod-
uct-related bene-
Perceived prod-
uct-related risks
General attitudes
Perceived proc-
ess-related risks
Perceived proc-
ess-related bene-
Attitude towards
product
Attitude towards
GE in food pro-
Attitude towards
process
Empirically
determined
standardised
regression
coefficient
weighted be-
liefs or belief
factors
Predictive Analysis
Influencing Factors
The next question to be answered was concerned with influencing factors of attitude. Factors which were believed to have been important determinants of
attitude had been previously alluded to in prior studies. Much of the work in this area had been based on two models, Ajzens Theory of Planned Behaviour
and Fishbeins Muliple-attiribute attitude model. Modifications were made to these models by researchers over the years, such as the inclusion of new parame-
ters, in order to make them more applicable to the field of GE. These new models based on theory of planned behaviour and multiple-attribute attitude model
are depicted above left and right respectively. In his 2001 study, Small had indicated factors he considered important and created a model (Below), based on
these previous studies. The first step in answering the second question involved the attempted recreation of Smalls model. The purpose of this exercise was
to test the validity and fit of this preconceived model for the 2009 data. This required the replication of the constructs which form the 2005 model. The va-
lidity and fit of the 2005 model for the 2009 data set was determined through regression analysis where R-values were calculated.
The next step involved attempted optimisation of the model for the 2009 data. Since the questionnaire included multiple groups of redundant items, which
measured the various construct but were not included in the final model as they werent perceived to hold much influence, it would be possible to test the
strength of the influence of these items in the 2009 data. All items from each construct group were run through their own regression model and the -
values of each item were observed. Items which were perceived to have the strongest influences, by observing -values, were selected to be included in the
new optimised model. The constructs of Smalls model were then created through summation of the selected items. The newly formulated constructs were
then run through a regression model and the R-value of the calculated. The R-value of the new model was compared with that of the 2005 model to deter-
mine if the new model had a better fit. If the new model did appear to be a better fit, then it may be suggested that the newly selected items are better
measures of the constructs and are thus better determinants of attitudes which influence behavioural intention.
Reliability and internal consistency of the selected items was also measured assessed. This was done through the calculation of Cronbachs Alpha in conjunc-
tion with correlation analyses.
Past Research
This thesis is the fourth instalment in a series of studies conducted by Bruce Small over the past 8 years. As with the previous studies, in 2001, 2003 and 2005, a
questionnaire was used to obtain information from the public with regards to their attitudes towards GE. It is these studies which have be used as a point of com-
parison to determine if there have been changes in the NZ publics attitudes towards GE.
The Issues & The Potential
The area of GE has, for a long time, been an area of hot debate often evoking
strong emotional responses from the public. Opponents to the technology ar-
gue that GE poses as a threat to the health, safety and wellbeing of the environ-
ment, animals and humans alike. Furthermore, GE has struck a chord with those
with a religious disposition as many argue that the very act of modifying an or-
ganisms DNA is against God/nature, playing God and outright unnatural.
Rebuttals to these arguments, from proponents of the technology, is centred
around the great potential of the technology to solve a number of issues cur-
rently plaguing the world. The use of GE in plants and animals used for food
could provide a solution to the global food crisis. The application of GE in medi-
cine could address the increasing incidence and prevalence of disease world-
wide, and finally, the application of GE in biofuels may provide reprieve for the
global fuel crisis.
The GE debate is an intrinsic and extrinsic moral minefield. In order to deter-
mine currents attitudes both deontological and teleological views must be
taken into consideration.
Thesis Title
Demystifying the New Zealand Publics Attitudes Towards Genetic Engineering: An Empirical Analysis.
The Team
The team involved in this project includes Blaine Ah Yuk-Winters and Rick Starr from the University of Auckland, together with Bruce Small and Dr Vish
Vishwanath from AgResearch.
The Project
The aim of the project is three fold. Firstly, to determine what the current attitudes of the NZ public towards GE are, particularly with a focus on its ap-
plication in foods. Secondly, to resolve what factors influence these attitudes and lastly, to establish if there were any changes in attitudes from previous
studies. This will all be achieved through the analysis of empirical data obtained via the use of a national survey of a randomly selected sample of the New
Zealand population.
The proposed outcome of this project will be insight into how the NZ public view GE as well as the validation of an existing model or the development
of a new model which show the factors which influence these attitudes and their relationship. This would be of value as it suggests areas or applications of
GE the public will be more accepting of as well as suggest potential strategies to help improve public attitudes and relations, towards and with, GE.
The Process
A questionnaire, based on the surveys conducted in previous years, was created by Mr
Small. With the aid of a marketing firm, the survey was distributed and data collected. This
data (n = 1008) was then collated and formatted for the analysis steps. In order to ascer-
tain answers to the research questions posed, the following process of data analysis was
undertaken.
Descriptive Analysis
Descriptive analysis was employed in order to ascertain information regarding the current
attitudes of New Zealanders. This involved the calculation of the means, standard devia-
tions, and frequencies of the Likert scores of the various items measured in the question-
naire.
Results
Descriptives and Frequencies
Due to the design of the Likert Scale, means >2.5 are considered to be negative
attitudes towards GE , whilst scores