*4.4. Variable Evaluation*

The models were developed according to the types of timber joints (five in total), and each model was composed of four visual appearance components. Each component was evaluated to ascertain which of the variables contributed to the prediction of overall aesthetic preference. Table 3 illustrates information about the independent variable and how it affects the dependent variable.


**Table 3.** Coefficients (a) of aesthetic preference of timber joints.

<sup>a</sup> Dependent variable: overall aesthetic preference.

#### *4.5. Significance*

The following independent variables have a statistically significant impact on the outcome variable (overall aesthetic preference) according to joint type:

Joint 1: The joint appears innovative


It is evident from the above that "The joint appears innovative" has a statistically significant impact (*p* < 0.001) on overall aesthetic preference for all joint types.

#### *4.6. Unstandardised Coefficients*

Unstandardised coefficients (B) indicate how much the dependent variable varies with an independent variable when all other independent variables are held constant. Since 'The joint appears innovative' is statistically significant across all joint types, we look at its influence on overall aesthetic preference. Table 3 also indicates that if 'The joint

appears innovative' index increases by a value of 1, we observe 0.347, 0.408, 0.347, 0.330, and 0.483 units increase on the dependent variable for Joint 1 to 5, respectively. So, the more a respondent perceives the timber joint to look innovative, the joint becomes more aesthetically pleasing and is hence preferred. Clearly, the effect of this independent variable is more pronounced on Joint 5. However, the 'confidence interval' for Joint 5 indicates that there is a 95 percent chance that the actual value of the unstandardised coefficient is between 0.291 and 0.675.

### *4.7. Standardised Coefficients*

Table 3 illustrates the contribution of each independent variable included in the model contributed to the prediction of the dependent variable. The beta value in this table is the standardised coefficient. These values for each of the different variables have been converted to the same scale for comparison. Therefore, the higher the beta value, the stronger the unique contribution to explaining the dependent variable. For Joint 1, as seen in Table 3, 'The joint appears innovative' had the largest beta coefficient of 0.39. Therefore, 'The joint appears innovative' caused the strongest unique contribution to explaining 'Overall aesthetic preference' when all other variables in the model were controlled. This was followed by 'The joint appears strong' (0.21), 'The joint appears long lasting' (0.13), and 'The joint appears functional' (0.07) and made the least contribution to predicting overall aesthetic preference. From Table 3, Joints 2, 3, and 5 can be explained in the same fashion where 'The joint appears innovative' had the largest beta coefficient. However, for Joint 4, we see that 'The joint appears functional' had the largest beta coefficient (0.39), followed by 'The joint appears innovative' (0.31).

Further potential information generated from Table 3 is the part correlation coefficient. It shows how much of the total variance in the independent variable is uniquely explained by that variable, and how much R square would drop if it was not included in the model. For Joint 1, 'The joint appears innovative' had a part correlation value of 0.308. This value was squared and multiplied by 100 to ascertain the percentage of variance [41]. The new value came out as 9.48. This represented that the component uniquely explained 9.5% of the variance in overall aesthetic preference (Table 4). Table 4 also explains which of the variables included in the models contribute more to the prediction of overall aesthetic preference for other joints. For example, 'The joint appears innovative' makes the strongest unique contribution to explain the overall aesthetic preference for Joints 2, 3, and 5. It explains 6.4%, 9.5%, and 11.1% of the variance in overall aesthetic preference for Joints 2, 3, and 5, respectively. Joint 4 illustrates a slightly different scenario. 'The joint appears functional' makes the strongest unique contribution to explaining 5.6% of the variance in overall aesthetic preference. However, 'The joint appears innovative' trailed by just 1%. 'The joint appears strong' explained variance in overall aesthetic preference for Joints 1, 3, and 5. However, Joint 5 (2.1%) was only statistically significant. The only other statistically significant variable is 'The joint appears to be long lasting', which explained a 4.12% variance in overall aesthetic preference for Joint 2.

Table 3 also shows two exceptional cases where for Joint 2, a 1 unit increase in 'The joint appears strong' is associated with a 0.078 unit decrease in 'overall aesthetic preference' and Joint 5, a 1 unit increase in 'The joint appear long-lasting' is associated with a 0.031 unit decrease in 'overall aesthetic preference'. However, both cases were statistically insignificant and omitted from further investigation.

**Table 4.** Percentage of variance explained by visual appearance components in overall aesthetic preference for the timber joints.

\* *p* < 0.001
