New information and communication technologies (ICTs) have become particularly widespread and are increasingly utilized in modernized cultures. Due to their frequent use and their ubiquitous and ever-present nature, ICTs are often perceived as an irreplaceable part of a highly dynamic and interconnected society [1]. In particular, smartphones, as the latest evolution of mobile ICTs [2], have signaled the start of the mobile age. According to the 2015 Pew Research Center survey, in 2015 approximately two-thirds of Americans had a smartphone, and this figure is expected to grow in the coming years [3]. In Italy, according to the 2015 survey carried out by the National Institute of Statistics, 54% of families had a smartphone in 2014, which represents an upward trend with respect to 2013 [4].

However, although mobile devices enable users to perform a variety of tasks in an unprecedentedly rapid, easy, and effective way, they can also lead to serious medical problems. These problems include exposure to radiation, “screen dermatitis,” tumors, and infertility [5,6]; mobile devices can also interfere with driving safety and cause accidents [7,8]. Furthermore, the pervasiveness and intrusiveness of such devices could lead to mental health issues such as problematic use, distress, and compulsive usage, culminating in what is termed as “smartphone dependence” or “smartphone addiction” [9]. Some scholars consider this behavioral addiction to be a variant of technology addiction (or technopathy), while others consider it to be a specific addiction. Experts describe smartphone-related psychological disorders and syndromes, such as “short message service texting addiction,” compulsive selfie-taking behavior, sexting, or phubbing, among others [10]. Sexting can be defined as a behavior consisting of sending and receiving sexually explicit messages [11], whereas phubbing (snubbing someone in a social setting in favor of one’s own mobile phone) may lead to the impairment of social life and couple relationships [12].

Nomophobia, a neologism that is derived from the combination of “no mobile,” “phone,” and “phobia” has recently emerged as a modern problem, denoting the fear of feeling disconnected. Nomophobia is currently considered a situational phobia [13]. In more details, nomophobia can be characterized as:

Symptoms that characterize nomophobia include an excessive use of a mobile phone, which is kept permanently switched on, with the subsequent feeling of anxiety at the thought of a lack of network coverage. Other symptoms include the habit of continuously looking at the mobile screen in order to check for messages or missed calls (ringxiety, a combination of the words “ring” and “anxiety”), and the false sensation of hearing a mobile ring or vibration (the so-called “phantom vibration syndrome”) [13]. To the best of our knowledge, no psychometric scales are available in the Italian language for investigating this psychological construct. We therefore planned a translation and validation study of the Nomophobia Questionnaire (NMP-Q), which is an instrument developed by Yildirim and Correia [2].

A total of 403 subjects (average age 27.91 years, SD 8.63; 160 males and 243 females, representing 39.7% and 60.3% of the sample, respectively) volunteered to take part in the study and were administered the Italian version of the NMP-Q (Textbox 1).

Regarding the schooling level, 2 subjects (2/403, 0.5%) had completed only elementary school, while 39 and 173 had completed middle school (39/403, 9.7%) and high school (173/403, 42.9%), respectively. All other subjects (189/403, 46.9%) had received higher education.

Concerning average smartphone use, 45 subjects usually spent less than 1 hour on their mobile phone (45/403, 11.2%), 94 spent between 1 and 2 hours (94/403, 23.3%), 69 spent between 2 and 3 hours (69/403, 17.1%), 58 spent between 3 and 4 hours (58/403, 14.4%), 48 spent between 4 and 5 hours (48/403, 11.9%), 29 spent between 5 and 7 hours (29/403, 7.2%), 36 spent between 7 and 9 hours (36/403, 8.9%) and 24 spent more than 10 hours (24/403, 6.0%). The means, SDs, and skewness and kurtosis figures are reported in Table 1. This information confirmed the normality of data distribution.

Since 403 subjects constituted a sample size large enough to compute reliable estimations of correlations among variables, we proceeded with the EFA. The N/k ratio (number of subjects/number of items) was 20.15:1, thus satisfying the requirements for commencing a factor analysis. Barlett’s test of sphericity was significant (Chi-square=5,796.275, degrees of freedom=19, P<.001) and the KMO index was .941.

As a result of the initial solution, four factors explaining up to 65.90% of the total variance were extracted. Before varimax rotation was performed, the first factor explained 23.32% of the variance, while the second and third explained 23.91% and 18.67%, respectively. After rotation, the three factors explained 23.32%, 23.91%, and 18.67% of the variance, respectively. The loadings of all items on each factor are shown in Table 2. Table 3 provides a summary of the EFA results and the reliability analysis of all items.

As seen in Table 2, the results of the second PCA run showed that each item loaded on a single factor, and that there was no cross-loading on other factors except for Item 5, which had a loading value of .452 on D1 and .520 on D3. This item was considered to load on its primary factor because this loading was more salient, thus explaining more variance, and was more theoretically reasonable. Furthermore, this value reached the cut-off value of .45; therefore, its loading on the secondary factor was not considered.

The Italian version of the NMP-Q showed an overall Cronbach alpha coefficient of .95 (.94, .89, and .88 for the three factors). As such, the internal consistency of the questionnaire can be considered good to excellent. The effect of dropping each variable per time is shown in Table 3. Removing each item per time leads to a decreased alpha coefficient for each item, showing that all items are essential to guarantee the questionnaire’s reliability.

The NMP-Q total score correlated with the number of hours spent using a mobile phone (standardized beta-coefficient=.385, P<.001), and with gender (borderline statistical significance: standardized beta-coefficient=.091, P=.057), whereas no statistically significant associations could be detected with age and schooling level (Table 4). Similar patterns were found for the total score of the NMP-Q subscale “not being able to access information,” with statistically significant associations with the number of hours spent on a smartphone (standardized beta-coefficient=.362, P<.001), gender (borderline statistical significance: standardized beta-coefficient=.088, P=.067) and schooling level (standardized beta-coefficient=.108, P=.024). The total score of the NMP-Q subscale “giving up convenience/losing connectedness” showed correlation only with the number of hours spent on a smartphone (standardized beta-coefficient=.402, P<.001). Finally, the score of the last NMP-Q subscale “not being able to communicate” correlated with the number of hours spent on a mobile device (standardized beta-coefficient=.254, P<.001) and with gender (standardized beta-coefficient=.142, P=.004).

In summary, the number of hours spent on a mobile phone turned out to be a predictor of all subscales, while gender was associated with D1 (although borderline significant) and D3. Finally, the schooling level correlated with D1. Further details are reported in Table 4.

The impact on each item of age, gender, schooling level, and number of hours spent using a mobile phone is shown in Multimedia Appendix 1. Age had an effect on items 9 (borderline statistical significance: P=.063), 10 (borderline statistical significance: P=.089), and 15 (P=.003). The number of hours spent on one’s own mobile device affected almost all items (1-4, 6-10, 15-20), whilst for items 5, 11, and 12 a borderline statistical significance could be found, and items 13 and 14 had no significance. Gender impacted only items 8 and 11 in a statistically borderline way (P=.087 and P=.075, respectively). Finally, schooling level affected items 1 (P=.098), 2 (P=.074), 5 (P=.086), 15 (P=.085), and 19 (P=.086), whereas this parameter had a significant impact on items 9 (P=.021), 10 (P=.004), and 13 (P=.037).

The interaction between the number of hours spent on a mobile device and gender significantly affected only item 10 (P=.016), whereas this parameter resulted in statistically borderline results for items 2 (P=.095), 7 (P=.080), and 19 (P=.074). The interaction between the number of hours spent on one’s own smartphone and schooling level resulted in significance for items 9 (P=.003), 10 (P=.044), 11 (P=.019), and 18 (P=.010), and showed statistically borderline significance for items 2 (P=.073), 12 (P=.057), and 17 (P=.056). The interaction between schooling level and gender significantly impacted items 2-5. Finally, the interaction between the number of hours spent on a mobile phone, gender, and schooling level was statistically significant for items 3 (P=.011), 10 (P=.045), and 13 (P=.041), whereas they resulted in borderline significance for items 4 (P=.067), 16 (P=.066), and 20 (P=.053).