**8. Methodology of data analysis and results**

46 Multivariate Analysis in Management, Engineering and the Sciences

crisis in the twenty-first century.

**Table 4.** Proposed Indicators for Attitudes toward technological development.

human contamination.

hunger.

**7.4. Sampling and data collection** 

Source: [10]

choices.

population ([93, 94]).

Administration and Education.

which has more features and functions.

ATI 11: I am aware that my technological choices will help to overcome the water

ATI 12: Having financial conditions, when buying a new phone, I choose the one

ATI 13: With the safe use of technology it is possible to protect nature from

ATI 14: avoid using technological artifacts that cause environmental destruction. ATI 15: I know that genetically modified foods may be the solution to world

ATI 17: I admit the exploitation of nature instead of the welfare of humanity.

ATI 16: I do not buy furniture that is not made from certified wood.

In this research we adopted the technique of cross-section as it brings the advantage of allowing the acquisition of a picture of the variables of interest at a given moment in time and to emphasize the selection of a significant and representative sample of the target

The four institutions that represented the sampling unit were selected considering the criteria of being institutions both public and private. The selected public university, located in Campinas/SP, has students from different regions of Sao Paulo State, as well as the other three private institutions. These private institutions were one university and one faculty of Sao Paulo/SP and one faculty of Campinas/SP. The two private faculties selected receive students from different regions in the state and were also chosen because the researcher had already served for a long period in one of them and is now starting activities in the other one. The diversity of courses that the four institutions have was also a decisive factor in their

The data collection in the public institution was done directly with the students, from different courses, and the questionnaires were, in the most part, passed before the beginning of the classes in the days chosen for the data acquisition. Students were selected from the following courses: Environmental Engineering, Computer Science, Nutrition, Psychology, Business Administration with emphasis in International Business, Electrical Engineering, Production Engineering, Physics, Mathematics, Technology in Environmental Management,

Initially, around 1006 questionnaires were returned, yielding a proportion of almost 23 interviewed by assertion. However, LISREL software was used in a procedure that made the disposal of questionnaires that were not fully answered. Thus, the amount passed to 600 valid questionnaires, representing a proportion of nearly 14 respondents per statement, which is a significant value considering [90] as basis, and taking into account that the model

is not complete and it still gave a good fit in LISREL software.

Following guidelines from [90], at the end of the collect, the data recorded in the questionnaires were entered in an Excel spreadsheet to be later processed by specific statistical software's to aid in the treatment and analysis of quantitative data. The software *SPSS* 13.0 was used to verify the reliability and constructs unidimensionality, as well as the system *LISREL 8.54,* one of the most traditional statistical structural equation modeling package that became popular in social science research, as shown in [95], and has adequate resources to the purposes of this research ([91, 95-102]). A The coding was made with the SIMPLIS command language, available in the system, which made possible the estimation of the parameters of the model through confirmatory factorial analysis, according to different estimation methods, and the verification of the respective measures of adjustment of the models.

#### **8.1. Individual evaluation of the constructs**

From the individual evaluation of each construct was then possible to conduct the validation of the models of measures of each of these (DSO, ATI and CON) and this validation was performed by applying the Confirmatory Factorial Analysis (*Confirmatory Factor Analysis - CFA*). This technique has the purpose to test the hypothesis of adjustment of empirical data to a theoretical model, where a relationship structure is imposed and confirmed by analysis. Nevertheless, the variables need not to be related to all common factors. In particular, as is the case of this investigation, each variable is related to only one factor.

#### **8.2. Unidimensionality of the constructs**

The constructs presented earlier had their dimensionalities tested since this action is an premise to the reliability of the construct. The observation of the unidimensionality was made observing if each value of the normalized residue matrix of the construct was lower than 2.58, in modulus, at a level of significance of 1%, indicating if the effect on the overall adjustment of the model was low. In each process the indices of fit were checked, supplemented by information generated by the option "Modification Index" programmed in LISREL ®, which points out how much is expected to decrease the chi-square if a given a reestimation occurred, as in [98]. A detailed analysis of the standardized residuals of all dimensions was made and it was found that the overall quantity of residues which exceeds the value of 2.58 is very low and don't reaches 3% of the total. Thus, the unidimensionality of the constructs is not compromised.

#### **8.3. Reliability of the constructs**

Reliability is a measure of the internal consistency of the construct indicators and of the adequacy of the scales to measure it. According to the authors, a value commonly used for acceptance of reliability is 0.70, although this is not an absolute standard, and values below

0.70 have been accepted if the research is exploratory in its nature and this value was observed in the research. The results from each one of the dimensions are indicated in the following table (Table 5):

Technology and Society Public Perception: A Structural Equation

Modeling Study of the Brazilian Undergraduate Students' Opinions and Attitudes from Sao Paulo State 49

**structural equations** *t-values* **R2**

CON = 1.016\*DSO 6.220 1.033

DSO = 0.880\*CON 1.288 0.774

CON = 1.069\*ATI 6.348 1.144

These measures were used as a way to evaluate each construct and the integrated model, because an adjusted model provides a benchmark for the confirmation of the validity of the constructs and the relationships among them, with respect to the complete structural model.

Several indicators were excluded in an attempt to get the best fitted model resulting in a total of 44, 19 indicators on the scale validated following the guidelines of [90]. Applying the MLE technique to estimate the model with antecedents in the social dimension, we obtained the structural equations, t-values of the estimated parameters and their respective R2, as shown in Table 7 for the estimation of DSO1, the t-values are above to 1.96 for a level of significance of 5%. This demonstrates the significant contribution of the endogenous constructs (conceptions and attitudes) for the Social Dimension (DSO) predictor construct and we have this model as the most adequate, satisfying the theory and our initial

**Models METHOD OF ESTIMATION MLE**

**Table 7.** Complete model estimated according to the MLE method.

**8.6. Presentation of the fitted model** 

**ATTITUDES TOWARDS THE TECHNOLOGICAL DEVELOPMENT** 

**Table 8.** Adjusted Model of Measure.

DSO1 ATI = 1.096\*DSO 7.708 1.202

DSO2 DSO = 0.116\*ATI + 0.764\*CON 0.188 e 0.906 0.795 CON1 ATI = 1.109\*CON 7.896 1.231

CON2 CON = 4.408\*ATI - 3.386\*DSO 0.271 e -0.208 1.908 ATI1 DSO = 0.889\*ATI 1.354 0.791

ATI2 ATI = - 0.145\*DSO + 1.249\*CON -0.151 e 1.327 1.257

These results indicate that the model which predicts the other variables is the DSO1.

From the observations presented in the previous sections, the best fitted model showed a number of constructs and indicators below the initial, as shown in Table 8 and graphically

**CONSTRUCTS VARIABLES**

**CONCEPTIONS OF TECHNOLOGY** CON 01; CON 03; CON 05; CON 07

**SOCIAL DIMENSIONS** DSO 01; DSO 02; DSO 04; DSO 06; DSO 08

ATI 01; ATI 05; ATI 06; ATI07; ATI010; ATI 11; ATI 12; ATI 13; ATI 15; ATI 16

**8.5. Evaluation of the integrated model** 

hypothesis.

Source: LISREL® Software

represented in Diagram 2.

Source: [10]


Source: Lisrel® Software

**Table 5.** Composite Reliability of the Constructs

As can be seen, the values are higher than the reference commonly established when calculated for each of the constructs. This indicates that the measures performed are suitable.

#### **8.4. Adjustment measures of the constructs**

In this step we evaluated all the models seeking to understand the structural relationships hypothesized. The most common procedure for the estimation of these parameters and which usually has higher efficiency, in accordance with [90], is the Maximum Likelihood method (*Maximum Likelihood Estimation – MLE).* The results achieved (Table 6) with the MLE method were well adjusted, considering the values given in the literature.


Source: [10]

**Table 6.** Comparison of the Measures of Adjustment of the Model with MLE.

These measures were used as a way to evaluate each construct and the integrated model, because an adjusted model provides a benchmark for the confirmation of the validity of the constructs and the relationships among them, with respect to the complete structural model.

#### **8.5. Evaluation of the integrated model**

48 Multivariate Analysis in Management, Engineering and the Sciences

**Table 5.** Composite Reliability of the Constructs

**8.4. Adjustment measures of the constructs** 

following table (Table 5):

Source: Lisrel® Software

**Main Indicators of the Adjustment of the** 

Weighted Chi-square

Root Mean Square

Normed Fit Index

Non-Normed Fit Index

Comparative Fit Index

Goodness of Fit Index

Adjusted Goodness of

Source: [10]

**Model** 

Error of Approximation (RMSEA)

0.70 have been accepted if the research is exploratory in its nature and this value was observed in the research. The results from each one of the dimensions are indicated in the

**Constructs Composite Reliability of the Construct** 

As can be seen, the values are higher than the reference commonly established when calculated for each of the constructs. This indicates that the measures performed are suitable.

In this step we evaluated all the models seeking to understand the structural relationships hypothesized. The most common procedure for the estimation of these parameters and which usually has higher efficiency, in accordance with [90], is the Maximum Likelihood method (*Maximum Likelihood Estimation – MLE).* The results achieved (Table 6) with the MLE

Degrees of freedom 144 143 144 143 144 143 X Chi-square 218.865 218.131 218.865 218.131 218.16 218.131 X

(χ**2/**GL) 1,52 1,53 1,52 1,52 1,52 1,53 lower than

(NFI) 0.817 0.818 0.817 0.818 0.818 0.818 Over than

(NNFI) 0.913 0.912 0.913 0.912 0.914 0.912 Over than

(CFI) 0.927 0.927 0.927 0.927 0.928 0.927 Over than

(GFI) 0.962 0.962 0.962 0.962 0.962 0.962 Over than

Fit Index (AGFI) 0.95 0.949 0.95 0.949 0.95 0.949 Over than

**Table 6.** Comparison of the Measures of Adjustment of the Model with MLE.

**Values Obtained with the MLE Method REF.** 

5,00

and 0,08

0,90

0,90

0,90

0,90

0,90

**DSO1 DSO2 CON1 CON2 ATI1 ATI2 VALUES** 

0.0308 0.0309 0.0308 0.0309 0.0306 0.0309 Between 0,05

DSO Models 0,704161 CON Models 0,703772 ATI Models 0,716902

method were well adjusted, considering the values given in the literature.

Several indicators were excluded in an attempt to get the best fitted model resulting in a total of 44, 19 indicators on the scale validated following the guidelines of [90]. Applying the MLE technique to estimate the model with antecedents in the social dimension, we obtained the structural equations, t-values of the estimated parameters and their respective R2, as shown in Table 7 for the estimation of DSO1, the t-values are above to 1.96 for a level of significance of 5%. This demonstrates the significant contribution of the endogenous constructs (conceptions and attitudes) for the Social Dimension (DSO) predictor construct and we have this model as the most adequate, satisfying the theory and our initial hypothesis.


Source: LISREL® Software

**Table 7.** Complete model estimated according to the MLE method.

These results indicate that the model which predicts the other variables is the DSO1.

#### **8.6. Presentation of the fitted model**

From the observations presented in the previous sections, the best fitted model showed a number of constructs and indicators below the initial, as shown in Table 8 and graphically represented in Diagram 2.


Source: [10]

**Table 8.** Adjusted Model of Measure.

Technology and Society Public Perception: A Structural Equation

Modeling Study of the Brazilian Undergraduate Students' Opinions and Attitudes from Sao Paulo State 51

showed that the undergraduate students surveyed gave evidence that the government, the research and educational institutions, and also the citizens in general, are the components that best represent, or could represent, the society in processes of technological decision making. Either for the choice of new technologies as for the development of other, with these points in common it is possible to say that undergraduates expect a government

Complementing the scenery, the indifference presented regarding the environmental nongovernmental institutions may not mean a lack of opinion, but tend to show that issues of sustainability and preservation of the planet must be taken into consideration. This will be

Summarizing these statements we can say that government, people, academic and educational sectors should join forces for better choices and technological decisions. This statement shows that all the literature indicated in Section 4 is consistent with the fitted model. Thus, these considerations show the reflex of the society in the attitudes of people in their technology choices and it can also be said that this reflex is also present in the way they

In the case of the conceptions, for one of the dependent constructs, we can infer that the undergraduates surveyed show three basic aspects that commonly appear as indicators of common sense of the interpretation of technology. More generally, one can say that the research has shown that technology is understood by the great majority of undergraduates as being intellectualistic and synonymous of science, as well as neutral and instrumentalist. This way, the research shows that understanding technology as a practical knowledge derived from scientific theoretical knowledge, or even mistakes it with science, is a strong indicator. This means that a deeper reflection about the production process of technology and all of its real reasons of conception are not present in the majority of respondents. The instrumentalist conception supports this conclusion significantly. If the process is not considered, there remains only the product. And the use of this product is sole responsibility of the person who acquires it, not reflecting, this way, the interests of the developers. This statement is supported by the neutral view of technology also present in

In general, even the concept of technology can be somewhat limited, its direct dependence on social factors is a good indicator and raises good evidence that undergraduates expect that influential sectors of the society act jointly in the development process of technology.

Regarding the attitudes (ATI), the second construct dependent on the social background (DSO), three points are possible to identify in a more comprehensive way: the awareness of the need to ensure the sustainability of the planet, consuming appropriate technologies, the use of technologies to socialize information and keep informed, and the practical manifestation of the instrumentalist conception. By taking again as a starting point the social dimension, we can infer that the students surveyed believe that jointly, government, teaching and research institutions and the population in general, can combine efforts to the pursuit of sustainable development. Thus, progress is possible if, and only if, political, social

position at the same time they feel able to participate in a more active form.

evident in the analysis of the attitudes that will be made in future paragraphs.

understand the technology.

the model.

Source: LISREL® Software.

**Diagram 2.**Path Diagram of the Integrated Model.

In summary, it can be said that the final model proposed was adequate and the various factors, in turn, significant. Thus, it is clear that the social dimension, measured by the DSO construct, can be considered a predictor of the attitudes (ATI) and conceptions (CON) related to the technology and that, by the unadjusted models, the inverse relationships are not true for the sample consulted.

#### **9. Concluding remarks**

Seeking an understanding of how elements of society can influence the conceptions and attitudes of individuals concerning the technological development, this research developed a theoretical model from which a research tool was elaborated and applied with undergraduate students.

It was found that all models were adjusted, but only the model DSO1 attested the research hypotheses according to the results previously presented.

The society, in the created model, was represented by different variables which represented specific sectors. Among all possibilities, in a comprehensive manner, the adjusted model showed that the undergraduate students surveyed gave evidence that the government, the research and educational institutions, and also the citizens in general, are the components that best represent, or could represent, the society in processes of technological decision making. Either for the choice of new technologies as for the development of other, with these points in common it is possible to say that undergraduates expect a government position at the same time they feel able to participate in a more active form.

50 Multivariate Analysis in Management, Engineering and the Sciences

Source: LISREL® Software.

**Diagram 2.**Path Diagram of the Integrated Model.

hypotheses according to the results previously presented.

not true for the sample consulted.

**9. Concluding remarks** 

undergraduate students.

In summary, it can be said that the final model proposed was adequate and the various factors, in turn, significant. Thus, it is clear that the social dimension, measured by the DSO construct, can be considered a predictor of the attitudes (ATI) and conceptions (CON) related to the technology and that, by the unadjusted models, the inverse relationships are

Seeking an understanding of how elements of society can influence the conceptions and attitudes of individuals concerning the technological development, this research developed a theoretical model from which a research tool was elaborated and applied with

It was found that all models were adjusted, but only the model DSO1 attested the research

The society, in the created model, was represented by different variables which represented specific sectors. Among all possibilities, in a comprehensive manner, the adjusted model Complementing the scenery, the indifference presented regarding the environmental nongovernmental institutions may not mean a lack of opinion, but tend to show that issues of sustainability and preservation of the planet must be taken into consideration. This will be evident in the analysis of the attitudes that will be made in future paragraphs.

Summarizing these statements we can say that government, people, academic and educational sectors should join forces for better choices and technological decisions. This statement shows that all the literature indicated in Section 4 is consistent with the fitted model. Thus, these considerations show the reflex of the society in the attitudes of people in their technology choices and it can also be said that this reflex is also present in the way they understand the technology.

In the case of the conceptions, for one of the dependent constructs, we can infer that the undergraduates surveyed show three basic aspects that commonly appear as indicators of common sense of the interpretation of technology. More generally, one can say that the research has shown that technology is understood by the great majority of undergraduates as being intellectualistic and synonymous of science, as well as neutral and instrumentalist.

This way, the research shows that understanding technology as a practical knowledge derived from scientific theoretical knowledge, or even mistakes it with science, is a strong indicator. This means that a deeper reflection about the production process of technology and all of its real reasons of conception are not present in the majority of respondents. The instrumentalist conception supports this conclusion significantly. If the process is not considered, there remains only the product. And the use of this product is sole responsibility of the person who acquires it, not reflecting, this way, the interests of the developers. This statement is supported by the neutral view of technology also present in the model.

In general, even the concept of technology can be somewhat limited, its direct dependence on social factors is a good indicator and raises good evidence that undergraduates expect that influential sectors of the society act jointly in the development process of technology.

Regarding the attitudes (ATI), the second construct dependent on the social background (DSO), three points are possible to identify in a more comprehensive way: the awareness of the need to ensure the sustainability of the planet, consuming appropriate technologies, the use of technologies to socialize information and keep informed, and the practical manifestation of the instrumentalist conception. By taking again as a starting point the social dimension, we can infer that the students surveyed believe that jointly, government, teaching and research institutions and the population in general, can combine efforts to the pursuit of sustainable development. Thus, progress is possible if, and only if, political, social

and economical efforts, are gathered in the search of a growth able to preserve the natural and material resources to ensure the well being of people.

Technology and Society Public Perception: A Structural Equation

Modeling Study of the Brazilian Undergraduate Students' Opinions and Attitudes from Sao Paulo State 53

*Centro Universitário Regional de Espírito Santo do Pinhal (UNIPINHAL), Esp. Sto do Pinhal, SP,* 

*Universidade Estadual Paulista "Júlio de Mesquita Filho", (UNESP), Guaratinguetá, SP, Brazil* 

[1] Cardoso, T. F. L. (2001). Sociedade e Desenvolvimento Tecnológico: Uma Abordagem Histórica. In: Grinspun, M.P.S.Z. (org.). Educação Tecnológica: Desafios e Perspectivas.

[2] Miranda, N. A. et al. (2007a). New Tchnologies of the Information and Communication in Education: A pre-test analysis In: 4th International Conference on Information Systems and Technology Management, 2007, São Paulo/SP. Anais do 4th CONTECSI.

[3] Miranda, N. A. et al. (2007b). Tecnologias da Informação e Comunicação na escola pública: realidade ou ilusão? In: II EIDE - Encontro Iberoamericano de Educação.

[4] Dagnino, R. (2007). Um Debate sobre a Tecnociência: neutralidade da ciência e determinismo tecnológico. Disponível em < http://www.ige.unicamp.br/site/aulas/138/ UM\_DEBATE\_SOBRE\_A\_TECNOCIENCIA\_DAGNINO.pdf >. Acesso em 8 Jan 2010. [5] Echeverria, J. (1998). Teletecnologías, espacios de interacción y valores. Teorema –

[6] Liguori, L. M. (1997). As Novas Tecnologias da Informação e da Comunicação no Campo dos Velhos Problemas e Desafios Educacionais. In.: LITWIN, E. (org.). Tecnologia Educacional: política, histórias e propostas. (Trad.: ROSA, E.). Artes

[7] Agazzi, E. (2002). El impacto de la tecnología. Biblioteca Digital da OEI. Disponível em: < http://www.argumentos.us.es/numero1/agazzi.htm >. Acesso em: 7 Jun 2009. [8] Berne, R. W. (2003). Ethics, Technology, and the Future: Na Intergenerational Experience in Engineering Education. Bulletin of Science, Technology & Society. Vol. 23.

[9] Gordillo, M. M. & Galbarte J. C. G. (2002). Reflexiones Sobre la Educación Tecnológica

[10] Veraszto, E. V. et al. Estudios CTS en Brasil: relación causal entre concepciones y actitudes de estudiantes universitarios del Estado de São Paulo frente al desarrollo

desde el Enfoque CTS. Revista Iberoamericana de Educación. No. 28: 17-59.

Estéfano Vizconde Veraszto and Dirceu da Silva

José Tarcísio Franco de Camargo

São Paulo. Cortez. p. 183-225.

São Paulo/SP. v.1. p.1590-1602.

Araraquara/SP. Anais do II EIDE.

Médicas, Porto Alegre. pp. 78-97.

N. 2. April 2003. pp. 88-94.

Revista Internacional de Filosofía. Vol.17/3.

Leandro Valim de Freitas

**10. References** 

*Brazil* 

*Universidade Estadual de Campinas (UNICAMP), Campinas/SP, Brazil* 

*Petróleo Brasileiro SA (Petrobrás), São José dos Campos, SP, Brazil* 

With a differentiated education, new educational public policies may be developed in order to point to a sustainable world, whose maintenance of the life forms and inanimate resources can only be achieved through the joint action of all sectors of society. With an efficient technological education it is possible to educate for the consumption in a conscious and not so materialistic way, as pointed by some attitudes of the model. With a conscious technological education and with the use of all socializing and educational potential of technology, it is possible to manage and generate gradually an educational system solid and participatory.

Finally, it is possible to establish a contrast with the idea of consumption that also appeared, which indicates that the economic sector of society, which aims at maintaining a competitive market structure, also exerts influence. Even so, the adjusted model showed that the attitudes (or intentions) of undergraduates show evidence of change of attitude in college students that, even in a subjective way, give evidence that it is necessary to create a social mechanism where the holders of the technical knowledge must meet with the representatives from all sectors of society to decide which new technological systems should be adopted, since it does not harm the environment.

We emphasize that it was found that undergraduate students surveyed have a limited conception of technology, and this, as showed in the survey, is a reflection of the society in which they are inserted. Another point to be considered is the social reflex in the attitudes of the individuals facing the technological development. The study showed that there is a sustainable awareness, but also showed that some key variables of technological advances presented in the research model does not appear so striking in the way the students surveyed are positioned. These two observations open up space for a last important point: society, in general, need a technological reeducation, so that the citizens within it start to understand the process of technological decision making in a more comprehensive way and become capable to reflect about the different aspects related to the social environment in which they belong.

Thus, speaking in rethink the public policies in education is to propose the use of technological knowledge in education. And that does not mean simply perform tasks for training or specialization in new technologies, but to ensure to the students a solid foundation that helps them to manage and generate, in the future, the demands placed on society. The integration means of the individual in society, as well as their formation, more critical and more human.
