*3.2. Model Results*

Table 5 reports the results of the estimation of the model in Equation (1). An χ<sup>2</sup> test on the joint significance of the coefficients of demography, education, and working status variables failed to reject the null hypothesis that all coefficients were jointly equal to zero (*p*-values 0.290 and 0.597 in the TNUT and TENV regressions, respectively) (The *p*-values of the χ<sup>2</sup> on the joint significance of specific groups of variables (demography, education, employment status) are reported in Table 3.)


**Table 5.** Regression results.

Note: Figures in Italic font report the *p*-values of a test on the joint significance of the regression coefficient of the corresponding group of variables (demography, education, working status, organic, origin).

Technophobia's coefficient was negative and statistically different from zero at a 95% confidence level in both regressions. High values of TFI were associated with a higher probability of being absolutely unwilling to try olive by-product food. Table 6 reports the marginal probabilities.

An χ<sup>2</sup> test failed to reject the hypothesis that ORGANIC did not affect TNUT. Instead, we detected a statistically significant association with TENV (95% confidence level). Holding all other variables constant, buyers of organic food are expected to be more willing to try food with olive by-products if the consumption is beneficial for the environment.

In addition, respondents considering product origin an important or very important issue in food choice are more likely to be very willing to try food with olive by-products. The results hold in both regressions.



202

#### *Foods* **2020** , *9*, 1325
