The mere presence of a smartphone reduces basal attentional performance

Statistical analyses

To test the hypothesis that subjects have lower attention performance when the smartphone is present, the d2-R concentration and attention test is conducted. The performance score of the d2-R represent the individual attention performance scores. The performance is evaluated for both conditions (with or without the smartphone). In addition, the German short version of the smartphone addiction scale is evaluated. Stronger and weaker tendencies towards smartphone addiction are examined. The results of the d2-R and the questionnaire are evaluated using SPSS Statistics for Mac. An effect could be shown.

A one-factor ANOVA was used to examine the subjects’ attention performance in connection with a tendency towards smartphone addiction. The presence or absence of the smartphone is the independent variable, and the attention performance score is the dependent variable. Group comparisons were made using the ANOVA. Additionally, a covariance analysis was conducted to picture the potential moderating effect of the covariate “smartphone addiction”.

Effects of smartphone presence on attention

For the hypothesis that people have lower attention performance when the smartphone is present, three one-tailed ANOVAs are performed with the presence or absence of the smartphone as the independent variable and one of the three given scores of the d2-R (attention performance, speed and error-value) as the dependent variable.

To perform an ANOVA, the following necessary assumptions for the validity of the results of an ANOVA are checked: the normal distribution of the scores, the homogeneity of the variance and the independence of the samples. There was no normal distribution given for all values. As an ANOVA is a robust test, an ANOVA was chosen as the used test. We use an alpha level of 0.05 for all statistical tests.

To test the hypothesis that there is a significant difference between the two groups, an ANOVA is performed with attention performance score (AP score) as the dependent variable and condition (with or without smartphone) as the independent variable. In addition, the effect sizes are given and are classified into small (η² = 0.01), medium (η² = 0.06), and large (η² = 0.14) effects. For the classification of the effect size η², reference values for classification according to Cohen are used.

The means of the AP, PTO and E% scores for the two conditions (with and without smartphone) are shown in Table 1. A significant effect was found (F(1,40) = 6.168, p = 0.017, η² = 0.134). The without smartphone group has significantly higher attention performance (M = 108.95) compared to the with smartphone group (M = 99.71). There is a tendency towards a large effect. In addition to the AP score as a dependent variable, the ANOVA is performed with the speed score (PTO score) as dependent variable and the condition (with or without smartphone) as independent variable. A significant effect is also found for this ANOVA (F(1,40) = 7.592, p = 0.009, η² = 0.160). The significance of this ANOVA means that the subjects under the smartphone condition (M = 98.48) have a significantly lower work speed compared to the without smartphone condition (M = 108.57). A large effect is present here.

To test the hypothesis that there is a difference between the without smartphone and with smartphone groups for accuracy (E% value), an ANOVA is performed with the E% score as the dependent variable and the condition (with or without smartphone) as the independent variable (F(1,40) = 2.088, p = 0.156, η² = 0.050). The result of this ANOVA is not significant, meaning that the subjects do not have significantly different scores in accuracy under the two conditions.

Thus, the ANOVA revealed that participants under the smartphone presence condition show significantly lower performance on the d2-R attention test compared to participants who complete the attention test in the absence of the smartphone. Both attention performance scores (AP scores) and working speed (PTO scores) are significantly lower in the smartphone presence group compared to the smartphone absence group. Only accuracy (E% score) does not differ significantly.

Moderation by smartphone dependence

The mean for the d-KV-SSS for the entire group is M = 29.47, SD = 8.38. The mean for the group of subjects in the with smartphone condition is M = 29.80, SD = 8.89. For the group in the without smartphone condition, the mean is M = 29.14, SD = 8.04. No significant mean difference is found, F(40,1) = 0.065, p > 0.05.

Thus, overall, no strong tendencies toward smartphone dependence can be found for the entire group of subjects. If the mean of the total group is considered, it is M = 29.47 and thus forms the midfield. This shows that there are no strong, but also no low tendencies towards smartphone dependence.

The possibility of smartphone dependence as a moderating variable is examined. For this purpose, covariance analysis is performed with the variable of smartphone presence or absence. This analysis shows no significant interaction of the variables. Thus, a moderation of the effect of the presence of the smartphone by the individual differences in the dependence on the own smartphone cannot be confirmed.