Content uploaded by Tiziano Gerosa
Author content
All content in this area was uploaded by Tiziano Gerosa on Oct 28, 2020
Content may be subject to copyright.
Running head: the Parental Phubbing Scale
1
“Mom, Dad, Look at me”:
The development of the Parental Phubbing Scale
Luca Pancani1*¶, Tiziano Gerosa2¶, Marco Gui2, & Paolo Riva1
1 Department of Psychology, University of Milano-Bicocca, Milano, Italy
2 Department of Sociology, University of Milano-Bicocca, Milano, Italy
* Corresponding author
¶ These authors contributed equally to this work
Authors’ Note
This work was supported by the University of Milano-Bicocca under the “Innovation Project
Grant 2016” (2016-CONV-0051). The authors would like to thank collaborators at all schools that
actively participated in data collection. Correspondence concerning this article should be addressed
to Luca Pancani, University of Milano-Bicocca, Department of Psychology, Piazza Ateneo Nuovo,
1, 20126 – Milano (Italy). E-mail: luca.pancani@unimib.it
Running head: the Parental Phubbing Scale
2
Abstract
The widespread diffusion of smartphones has opened new challenges regarding the psychological
consequences of their usage on social relationships. The term phubbing (a combination of phone and
snubbing) indicates the act of ignoring someone in a social context by paying attention to the
smartphone. The few existing studies show that phubbing is widespread, mutually reinforced, and
socially accepted, with possible negative consequences for social and individual well-being.
Phubbing can occur in every social context, including romantic relationships, workplaces, and family.
However, to date, minimal attention has been given to the possible impact that phubbing carried out
by parents can have on their children. To start filling this gap, in this paper, we introduced a new scale
that measures the perception of being subject to parental phubbing and showed the prevalence of
perceived phubbing on a stratified sample of 3,289 adolescents. Firstly, the dimensionality, validity,
and invariance of the construct were proven. Moreover, our results showed a positive relationship
between children’s perceived levels of parental phubbing and their feelings of social disconnection
with parents, thus suggesting that the more children felt that one or both of their parents were
phubbing them, the less the children felt connected with their parents.
Keywords: phubbing; smartphone; parenthood; social connections; scale development.
Running head: the Parental Phubbing Scale
3
Introduction
Imagine you are in a café. In a table in front of you, there is a father with his 13-year-old
daughter. The child talks excitedly to her father; however, his eyes are glued to the smartphone, and
his fingers keep scrolling down some discussion that is occurring online. Occasionally, he nods his
head and gives her minimal, inattentive replies at her questions. She raises her voice, touches his
arm, but all her attempts to draw his attention back to what she is saying seem worthless. How
would this young girl feel? How would her father feel while doing this? And how would you feel
observing this scene? This kind of situation should be quite familiar to most people nowadays, as it
has become increasingly common in contemporary daily life, to the extent that a neologism,
“phubbing,” was coined to refer to it (Macquarie, 2013). Phubbing is the combination of phone and
snubbing and can be defined as the act of ignoring someone in a social context by paying attention
to the smartphone. In this work, we introduce a new scale that measures the perception of being the
subject of parental phubbing.
Some recent studies showed how phubbing is widespread. In a survey conducted by
McDaniel and Coyne (2016), 70% of participants reported being phubbed by their partner,
especially during leisure time, in which phubbing situations seemed to occur at least once a day for
the 62% of the sample. These data are in line with those of Al-Saggaf and MacCulloch (2018), who
showed that 62.3% of their participants declared themselves as phubbers, reporting that the most
likely target of their phubbing behavior was their partner. However, these studies are limited by
their samples that are unbalanced in terms of gender (with a female prevalence), which potentially
creates a confounding effect between the prevalence of the phenomenon and gender differences.
According to Chotpitayasunondh and Douglas (2016), phubbing has become a sort of new social
norm, which was rapidly established through reciprocity, a fundamental process of human
interaction (Cialdini, 1993; Falk & Fischbacher, 2006). Indeed, ignoring someone because of the
smartphone might be mirrored by the ignored counterpart, intentionally or not. The authors showed
Running head: the Parental Phubbing Scale
4
that the relationship between phubbing and being phubbed is strong (β = .60) and that experiencing
phubbing (either actively or passively) increased the perception that this phenomenon is normative.
Thus, phubbing seems to be mutually reinforced in social interactions, and, according to the
authors, this might have led people to perceive this behavior as acceptable.
The State-of-the-Art of Research on Phubbing
Given the novelty of the phenomenon, little is known about the antecedents and
consequences of phubbing. However, the few studies that investigated possible determinants of
phubbing agree on one point: phubbing relates positively with “smartphone addiction”
(Chotpitayasunondh & Douglas, 2016; Karadağ et al., 2015) or, in other words, with excessive use
of smartphones that might lead to adverse effects on user’s daily life (King & Dong, 2017; Lee,
Chang, Lin, & Cheng, 2014; Lin et al., 2016). Although there is an ongoing debate about whether
smartphones are addictive (Gentile, Coyne, & Bricolo, 2013; Kardefelt-Winther et al., 2017), it is
perfectly understandable that individuals who use their smartphone longer and more frequently are
more likely to phub others.
However, research on phubbing has been primarily concerned about its consequences on
human relationships. The available studies seem to converge on the notion that phubbing causes
negative impacts. Indeed, being phubbed by the partner decreases relationship satisfaction that, in
turn, has an impact on depression and life satisfaction (Roberts & David, 2016). Similarly, a study
on married adults showed that partner phubbing behavior was indirectly associated with depression
by negatively impact on relationship satisfaction (Wang, Cie, Wang, Wang, & Lei, 2017).
Comparable findings emerged in research that investigated phubbing in a different context, namely
the relationship between supervisors and employees. Roberts, Williams, and David (2017) found
that being phubbed by their bosses negatively affected employees' engagement by decreasing their
trust in them; thus, similar to the romantic context, phubbing occurring in the workplace
undermines the quality of the relationships.
Running head: the Parental Phubbing Scale
5
Given the negative association between phubbing and quality of interpersonal relationships,
scholars have taken the first steps towards a deeper understanding of the phenomenon, investigating
it by adopting an experimental methodology. For instance, Abeele, Antheunis, and Schouten (2016)
found that people who used a smartphone during a conversation were perceived as less polite and
attentive than those who did not use it, especially when their phubbing behavior was self-initiated
and not done in response to a notification (i.e., smartphone vibration, sound, and lighting for an
incoming message). Even being only a witness of a phubbing scene, as in our first example, might
have some negative consequences for the observers’ mood, and it increases their stress level (Nuñez
et al., 2018).
In a recent study, Chotpitayasunondh and Douglas (2018) manipulated the phubbing
experience using a 3D animation of a conversation between two characters in which participants
were asked to imagine themselves as one of them. The longer the phubbing experience, the worse
the quality of communication and perceived relationship satisfaction. Moreover, phubbing intensity
negatively affected the satisfaction of the four fundamental needs as theorized by Williams
concerning the experiences of being ignored (2009): (1) belonging, that is the need to engage in
positive (or at least, not negative) interactions with other people; (2) self-esteem, which concerns
the need to maintain a positive view of ourselves; (3) meaningful existence, that is the necessity to
feel recognized by others and being worthy of attention; and (4) control, the need to perceive
influence over the surrounding social environment. Specifically, larger exposure to phubbing was
associated with a lower sense of belonging, a decrease in self-esteem, the perception of one’s own
existence as less meaningful, and less perceived control over the social environment. The strongest
impact of phubbing was observed for the need to belong, which was also responsible for the indirect
effect of phubbing on the quality of communication and relationship satisfaction. In keeping with
these findings, another recent study (Hales, Dvir, Wesselmann, Kruger, & Finkenauer, 2018) found
that participants, asked to recall a time in which a conversation partner used their cell phone during
Running head: the Parental Phubbing Scale
6
an interaction, felt ostracized (i.e., ignored). The authors also found that being phubbed resulted in
higher feelings of relational devaluation (i.e., perceiving themselves in the eyes of the partner as not
as important, valuable, or close as much as one desire; (Leary, 2001, 2020), which in turn accounted
for lower levels of basic needs satisfaction.
The latest research by Chotpitayasunondh and Douglas (2018) and Hales et al. (2018) is
particularly important because it emphasizes the link between the recently developed phenomenon of
phubbing and the long and fruitful research tradition on being and feeling excluded (for an overview,
see: Riva & Eck, 2016). Social exclusion has been broadly defined as the experience of being kept
apart from others physically (e.g., social isolation) or emotionally (e.g., being ignored or told one is
not wanted; Riva & Eck, 2016). In this perspective, ostracism (i.e., being primarily ignored) and
social rejection (i.e., being explicitly told one is not wanted) represent the two core experiences of
social exclusion. Since it involves being primarily ignored, phubbing represents an instance of
ostracism (Williams, 2007, 2009). Adverse effects of ostracism and rejection are well-known. They
include both short-term (e.g., negative emotions, antisocial behaviors, cognitive depletion) and long-
term (a series of negative mental and physical outcomes, including depression, poorer immune
functioning, and higher rates of substance use) aversive consequences (Bernstein, 2016; Riva,
Montali, Wirth, Curioni, & Williams, 2017).
Phubbing and Parenthood
Relational problems between parents and their children could arise from a wide range of
factors (e.g., quality of communication, parental warmth, attachment style, adolescents’
externalizing, and internalizing problems). Within this variety, technology might represent one of
the factors ubiquitously grabbing the parents’ attention (e.g., while at playing home, while eating,
while walking in the park), thus negatively affecting the parent-child relationship. Indeed, among
different technologies (e.g., TV, computer), the ubiquity of smartphones makes this device currently
the most likely source of parental distraction and deserves appropriate attention form scholars. The
Running head: the Parental Phubbing Scale
7
effects of relational devaluation linked with phubbing might be notably stronger when the
relationship at stake is relevant. For children, the most important source of meaning and social
support are their parents. Thus, the negative effect of phubbing might be even stronger in parent-
child relationships in which communication and parental responsiveness have central roles in
children and adolescents development (Baumrind, 1991; Caughlin & Malis, 2004; Davidov &
Grusec, 2006; Kochanska & Aksan, 2004; Pinquart, 2016). Indeed, McDaniel and Radesky (2018)
have recently shown that mothers' distraction with technological devices (what the authors called
“technoference”) is associated with problematic externalizing and internalizing behaviors of their
young children (5 years old or younger). In line with these findings, Stockdale, Coyne, and Padilla-
Walker (2018) showed that parental technoference was related to adolescents’ negative
psychological (i.e., higher anxiety and depression) and behavioral (i.e., cyberbullying) outcomes.
However, the authors also found positive associations with civic engagement (e.g., involvement in
political issues, time spent volunteering) and prosocial behaviors (i.e., helping family members and
strangers), which were interpreted as means to gain attention from parents who are distracted by
technological devices.
However, no instrument has been developed to date to assess parental phubbing. This is a
problematic omission, considering that the absence of a measuring instrument for parental phubbing
prevents the generation of basic knowledge about the diffusion of this phenomenon and its effects.
Thus, we argue that there is an urgent need for a psychometrically valid instrument to measure
adolescents’ perception of parental phubbing.
Existing Measures of Phubbing
In the last five years, scholars have proposed various measures of phubbing; each one
focused on a specific aspect of this phenomenon. To the best of our knowledge, the first self-report
scale on phubbing was developed by Karadağ and colleagues (2015). Based on focus groups, the
authors developed a 10-item scale that measured phubbing behavior by respondents towards other
Running head: the Parental Phubbing Scale
8
individuals. Specifically, the principal component analysis revealed two factors, both with good
internal consistency: five items measured communication disturbance (i.e., how frequently
respondents disturb ongoing face-to-face communications by using their smartphone) and the
remaining five items measured phone obsession (i.e., how much respondents need their smartphone
when they are not interacting with others). While the former factor strictly concerns phubbing,
phone obsession is closer to a general dimension of problematic smartphone use (Kwon, Kim, Cho,
& Yang, 2013; Pancani, Preti, & Riva, 2020) than a specific dimension of phubbing. Moreover, the
sample on which the analyses were conducted mainly consisted of females (71.6%).
Contrarily to the scale above, Roberts and David (2016) developed a measure of perceived
phubbing from others and, specifically, from the partner. The authors generated a large pool of
items that were reduced through expert evaluation, inter-rater agreement about face validity, and
exploratory factor analysis. This procedure led to retain nine items that loaded on a single, highly
reliable factor of perceived partner phubbing. The Partner Phubbing scale (Pphubbing) was then
used in other studies, both in its original form (e.g., Wang et al., 2017) and in a modified version
targeting phubbing on work setting (Roberts et al., 2017).
A completely different facet of phubbing has been captured by the measure proposed by
Chotpitayasunondh and Douglas (2016). The authors developed a 5-item scale to assess social
norms concerning phubbing, focusing both on descriptive (i.e., familiarity and spread of the
phenomenon) and injunctive (i.e., appropriateness of the behavior) norms.
Last but not least, two brief scales were conceived in the technoference literature, targeting
different technological devices in addition to the smartphone (e.g., tablet, computer, television).
McDaniel & Coyne (2016) developed the Technology Device Interference Scale (TDIS) and the
Technology Interference in Life Examples Scale (TILES). Both the scales aimed at measuring the
frequency of technology interference in romantic relationships. While the TDIS asked how often
each device disturbs or interrupts a dyadic interaction, the TILES asked how often each of five
Running head: the Parental Phubbing Scale
9
common situations of technoference (e.g., the partner pulls out the phone during mealtime or is
distracted by TV during a conversation) occurs. As evidenced by the items, the two scales were
developed to measure a phenomenon wider than phubbing, namely the overall interference of
technological devices on dyadic relationships. Moreover, two modified versions of the TILES were
developed to measure both adolescents’ technoference and perceived technoference from their
parents (Stockdale et al., 2018). However, besides internal consistency, no information about the
psychometric properties of the two scales was reported.
The Present Study
The present study aimed at developing a brief, psychometrically valid, scale to assess
parental phubbing in adolescents, the Parental Phubbing Scale (PPS). Specific aims were to (1)
develop items of the PPS and test its dimensionality, (2) investigate its measurement and structural
invariance across different subpopulations, (3) testing the PPS concurrent validity through its
association with children’s feeling of social disconnection, and (4) check differences in parental
phubbing levels due to a set of sociodemographic variables.
Several hypotheses were set. The PPS was developed to measure perceived phubbing
distinctly and separately from each parent, and we expected that the two dimensions would be
correlated and would jointly measure an overall dimension of parental phubbing. This theoretical
dimensionality was expected to be invariant across participants’ gender, ethnic origin (migrant vs.
native), mother, and father education level.
Consistently with the recent literature on phubbing (Chotpitayasunondh & Douglas, 2018;
Hales et al., 2018), we hypothesized a good concurrent validity of the PPS. Specifically, we
hypothesized that source-specific, perceived phubbing would be positively associated with
perceived social disconnection. In other words, we expected to find significant and positive
associations between participants’ perception of being phubbed by mother (father) and the feelings
Running head: the Parental Phubbing Scale
10
of disconnection from their mother (father), along with non-significant or, at least, lower cross
associations (i.e., being phubbed by mother and perceived disconnection from father and vice-
versa). Similarly, the overall parental phubbing would be significantly and positively associated
with the overall perception of social disconnection from parents.
Finally, concerning sociodemographic variables, we expected that a higher education level
might be related to lower phubbing habits. Indeed, research on the digital divide has found that
individuals with higher socioeconomic status show better digital skills and get higher benefits from
it (Van Dijk, 2005). Recent research also shows that those with a higher education level can better
cope with digital overuse (Gui & Büchi, 2019) and that parents with higher socioeconomic status
are more aware of digital overuse and more likely to impose restrictions on their children’s use of
digital media (Nikken & Opree, 2018). No hypotheses were advanced concerning gender and ethnic
origin.
Method
Participants and procedure
Participants were extracted from the second wave of a longitudinal survey, which is part of a
wider experimental project named “Digital Well-being - Schools” (Gui, Gerosa, Garavaglia, Petti,
& Fasoli, 2019), carried out by the authors. The data collection process was carried out in May 2018
and involved all the students in grade 10 (15-16 years old) enrolled in 18 high schools of two
neighboring school districts of the Lombardy region (Northern Italy). Students were surveyed
through a CAWI methodology (Computer Assisted Web Interviewing), asking them to fill in an
online questionnaire in the computer labs of their school under the supervision of external
observers. The questionnaire was finally administered to 3,289 participants located in 171 classes,
achieving a total response rate of 90%. A more detailed description of the sample characteristics is
offered in Table 1.
Running head: the Parental Phubbing Scale
11
[Table 1 near here]
Materials
The original questionnaire provided detailed information on students’ sociodemographic
characteristics, digital competence, attitudes toward digital technologies, and smartphone daily
usage habits (for complete information, see Gui et al., 2019). For the purpose of this study, only the
following measures were used.
Item Pool for the PPS
Two subscales of the PPS were developed: the PPS-Mother (PPS-M) and the PPS-Father
(PPS-F). The PPS-M and PPS-F were identical (i.e., included the same items) except for the source
of phubbing mentioned, mother and father, respectively. The items were adapted from those
included in the Pphubbing scale (Roberts & David, 2016). Specifically, the term “cell phone” was
replaced by “smartphone,” and the term “partner” was replaced by “mother” or “father,” according
to the subscale. No other changes were made for items 1, 2, 3, 4, and 9 of the Pphubbing scale.
Conversely, slight modifications were introduced for items 5 and 6. Indeed, item 5 of the
Pphubbing scale (i.e., “My partner glances at his/her cell phone when talking to me”) was modified
into “My mother/father get distracted when we do something together.” This change made the item
representative of different activities (i.e., not only talking) shared by adolescents and parents and
different ways in which phubbing can be put in place (e.g., glancing at the smartphone, phone calls,
playing with online gaming). Item 6 of the Pphubbing scale (i.e., “During leisure time that my
partner and I are able to spend together, my partner uses his/her cell phone”) was changed into
“During leisure time that we spend together, my mother/father pays more attention to her/his
smartphone than to me,” to detect more extreme situations. The two remaining items of the
Pphubbing scale were not included because item 7 was reverse coded, and it could generate
confusion, whereas item 8 reports a situation (i.e., going out together) that is more typical of
romantic relationships than adolescent-parent ones. Thus, each of the subscales consisted of 7 items.
Running head: the Parental Phubbing Scale
12
Regarding response options, a five-point Likert scale was used, ranging from 1 “Never” to 5 “All
the time.” The items were then translated into Italian and then back-translated into English by a
native English language translator to check their meaning was maintained.
Social Disconnection
We used three items to measure feelings of social disconnection towards parents.
Specifically, for each parent, participants were asked how often they felt (1) lack of companionship
from, (2) ignored by, and (3) left out. A five-point Likert scale was used, ranging from 1 “Never” to
5 “All the time.”
Sociodemographic Variables
Participants were asked to indicate their gender, ethnic origin, and the education level of
their parents. Students’ gender and ethnic origin were collected as dichotomous variables,
distinguishing males from females and natives from the first and second generations of migrants,
respectively. The level of education achieved by both parents was recorded in three reference
categories identifying low-educated (up to middle school diploma), middle-educated (up to high
school diploma), and highly-educated subjects (bachelor’s degree or higher).
Data analysis
All the analyses were carried out using Mplus, version 7 (Muthén & Muthén, 2015). Given
the nested nature of our data (i.e., adolescents clustered within classes and classes clustered within
schools), all the models were tested adopting a multilevel approach. Specifically, the Mplus’
analysis type “two-level complex” allowed us to estimate the hypothesized associations among
variables at the within-subject level by adjusting standard errors and chi-square statistics for both
adolescents’ class and school membership, The only exception was measurement invariance, which
could not be tested using the “two-level complex” approach with two clustering variables and the
grouping variables (i.e., gender, ethnic origin, parents’ educational background) at the within level
Running head: the Parental Phubbing Scale
13
(Kim, Kwok, & Yoon, 2012). Thus, we opted for a design-based approach (Kim et al., 2012;
“complex” analysis type), which allows for invariance testing, adjusting the chi-square statistic and
standard errors (Asparouhov, 2006). Although the design-based approach allows a single clustering
variable (adolescents’ class membership was chosen as more meaningful than school membership),
it represents a good alternative to multilevel SEM, showing equivalent performance when the
between- and within-model structures are assumed as identical, as it was in our case (Wu & Kwok,
2012). Both type “two-level complex” and “complex” were run using the MLR estimator which is
robust to data non-normality.
Data were analyzed in four steps. First, the factor structure of the PPS was investigated
using multilevel exploratory (EFA) and confirmatory (CFA) factor analysis. Initially, the subscale
referring to father (PPS-F) was randomly chosen to explore the factor structure of the scale through
a multilevel EFA. Then, the emerging structure was tested on the items referring to mother (PPS-M)
through a multilevel CFA. Once the dimensionality of the subscales was confirmed, a multilevel
CFA was used to test the hypothesized factor structure of the PPS on all the items by estimating two
first-order factors of parent-specific phubbing (regarding mother and father, separately) that loaded
on an overall second-order factor of parental phubbing.
Second, the measurement and structural invariance of the PPS was tested on groups of
students defined by gender, ethnic origins, and parental education. A series of five hierarchically
nested models were run for each of the sociodemographic variables, using multi-group confirmatory
factor analysis (MG-CFA). Each model tested a higher level of invariance by adding a set of
specific equality constraints across groups (Byrne, 1988; Meredith, 1993; Widaman & Reise, 1997).
More technical details on factor validity and measurement and structural invariance are given in the
Supplementary Material.
Third, the concurrent validity of the PPS was assessed using multilevel structural equation
modeling (SEM). Specifically, we tested the association between parental phubbing and children’s
Running head: the Parental Phubbing Scale
14
perception of social disconnection from parents, both from a source-specific (i.e., constructs related
to each parent) and general (i.e., overall measures of parental phubbing and parental disconnection)
standpoints.
Finally, multiple indicators and multiple causes models (MIMIC; Joreskog & Goldberger,
1975) were run to investigate the relationships between the sociodemographic variables and the
source-specific and overall dimensions of parental phubbing.
The goodness of fit of the models was evaluated according to the following indices: (a) the
chi-square statistic (χ²), (b) the comparative fit index (CFI), (c) the Tucker–Lewis index (TLI), (d)
the root mean squared error of approximation (RMSEA), and (e) the standardized root mean square
residual (SRMR). A model adequately explains the data when the χ² probability is lower than .05,
the CFI and TLI are higher than .90 (better if higher than .95), the RMSEA is lower than .08 (better
if lower than .05), and the SRMR is lower than .08 (Brown, 2015; Kline, 2015).
Competing nested models were generally compared using the chi-square difference test
(Kline, 2015): a significant probability (p < .05) associated to Δχ² means that the more restricted
model (i.e., less free parameters) fit the data significantly worse than the less restricted model (i.e.,
more free parameters); thus, the latter one should be preferred. As recommended by Cheung and
Rensvold (2002), in addition to the chi-square difference test, two further tests were used to
compare models in the measurement and structural invariance analysis: the McDonald’s centrality
index (Mc) and CFI difference. Values of ΔMc < 0.02 and ΔCFI < 0.01 are considered sufficient
clues of cross-group equivalence in the examined construct.
Results
PPS dimensionality
The dimensionality of the PPS was initially explored using multilevel EFA on the items
referring to father (i.e., PPS-F subscale). The first eigenvalue extracted (4.34) was much larger than
Running head: the Parental Phubbing Scale
15
the second one (0.71), clearly indicating the presence of a single factor. Loadings were generally
high, ranging between .61 (item 2) and .87 (item 5). The subscale unidimensionality was confirmed
through a multilevel CFA on the items referring to the mother (i.e., PPS-M subscale). Despite a
significant chi-square statistic, easy to reach with such a large sample size (Bentler & Bonnet,
1980), the multilevel CFA on the items referring to mother yielded good fit indices [χ²(35) =
602.30, p < .001; CFI = .961; TLI = .953; RMSEA = .071; SRMR = .029]. Standardized loadings
ranged from .69 to .85, confirming results obtained in the multilevel EFA.
Finally, the theoretical dimensionality of the PPS was tested using a further multilevel CFA
on all the items. Specifically, items of the two subscales were loaded on two first-order latent
factors, namely PPS-M and PPS-F, which represented the adolescents’ perception of being phubbed
by their mother and father, respectively. Moreover, an overall dimension of parental phubbing was
included in the model by estimating a second-order latent factor (i.e., PPS) on which the two first-
order factors were loaded. In addition, we estimated error covariances between item pairs that were
identical except for the source of phubbing.
Though the chi-square statistic was significant, the model, reported in Figure 1, yielded
excellent fit indices [χ²(160) = 1397.19, p < .001; CFI = .957; TLI = .951; RMSEA = .049; SRMR =
.032]. Standardized loadings on first-order factors ranged from .62 to .87 and the loadings of PPS-
M and PPS-F on the PPS general factor were both equal to .58. Error correlations were all
significant at p < .001 level and ranged from .15 to .30, except for the correlation between the pair
of items number 5, which was slightly lower but still significant, r = .08, p = .006.
[Figure 1 near here]
A further model was tested without the error correlations, but it yielded worse fit indices
compared to the above model [χ²(167) = 2452.14, p < .001; CFI = .921; TLI = .914; RMSEA =
.065; SRMR = .040]. Moreover, the MLR-corrected chi-square difference test indicated that the
Running head: the Parental Phubbing Scale
16
model with error covariances fitted the data significantly better than the one without error
covariances (Δχ² = 736.21, Δdf = 7, p < .001).
The internal consistency of the three latent factors was computed using Cronbach’s alpha.
Results indicated high reliability for all of them, with the coefficient of PPS-M (α = .91) slightly
higher than those of PPS-F and PPS (α = .89 for both).
Measurement and structural invariance
Instead of focusing on the overall PPS construct, measurement and structural invariance
tests have been conducted on the first-order model in which PPS-M and PPS-F were left free to
covary. Even though the two models can be considered substantively equivalent from a statistical
point of view, the first-order solution has the advantage of being immune to the risk of under-
identification in the multi-group analytical framework, simplifying the entire estimation process.
To check whether the factor structure of our model specification was consistent among
different sub-population of participants, a preliminary analysis of baseline models was conducted
on groups of students distinguished by gender (females and males), ethnic origins (natives and non-
natives), father and mother education level (low, middle, and highly educated). We found general
inflation on the χ² values, but RMSEA, CFI, and TLI alternative fit indices reached at least
acceptable values for all the groups of students we considered in the analysis (Table 2), indicating
that data was suitable to proceed with the measurement and structural invariance tests.
We then estimated a total of five hierarchically nested models for each of the grouping
variables. The first four models deal with measurement issues related to observable items, testing
configural, weak (or metric), strong (or scalar), and common residual covariance invariance,
respectively (Byrne, 1988; Meredith, 1993; Widaman & Reise, 1997).
Configural invariance represents the prerequisite condition for assessing the equivalence of
all the other parameters in the model and can only be satisfied if the construct at stake has the same
Running head: the Parental Phubbing Scale
17
number of factors and the same patterns of free and fixed factor loadings across groups (J. Wang &
Wang, 2012). This is the case of our model specification that showed to be at least acceptable in all
the cross-group comparisons according to the values of CFI, RMSEA, and TLI reported in Table 2.
We, therefore, proceeded with the analysis of weak measurement invariance, which has to
do with the equivalence of slope coefficients obtained regressing observed items on their underlying
latent factor (factor loadings). The models fitted the data well, and their values of ΔCFI and ΔMc
remained widely below the .01 and .02 cut-off thresholds suggested by Cheung and Rensvold
(2002), highlighting the tightness of the invariance hypothesis for all the sub-populations under
study.
Similar results can also be inferred from the analyses of strong and common residual
covariance invariance. Strong measurement invariance was evaluated constraining items intercepts
to be equivalent across groups, to check whether latent mean differences accounted for all the mean
differences in the shared variance of the observable items (Putnick & Bornstein, 2016). On the other
hand, invariance testing of residual covariances enabled us to assess whether the covariances
between residuals operate equally across different groups. Watching at the variations registered in
the CFI and Mc alternative fit indices, we can confirm the existence of full measurement invariance
between all the groups considered in the analysis.
[Table 2 near here]
In addition to the items related measurement invariance testing procedure, we finally
conducted a set of multi-group comparisons focused on PPS-M and PPS-F factor variances and
covariances, reflecting the structural equivalence of the derived latent constructs themselves
(Vandenberg & Lance, 2000). Also, in this case, the invariance hypothesis held for each of the sub-
population under study, both in terms of goodness of fit to the data and variations of the CFI and
Mc alternative fit indices with respect to the configural model.
Running head: the Parental Phubbing Scale
18
Association between phubbing and feelings of social disconnection
As a preliminary step, a second-order, multilevel CFA was run on the social disconnection
scale (SDS), confirming the hypothesized dimensionality with excellent fit indices [χ²(20) = 59.44,
p < .001; CFI = .995; TLI = .993; RMSEA = .025; SRMR = .023]. As for the PPS, we estimated
two first-order factors, one referring to mother (i.e., SD-M; α = .69) and one to father (i.e., SD-F; α
= .70), and a second-order factor (i.e., overall social disconnection, SD; α = .78).
Two multilevel structural equation models (SEMs) were tested to investigate whether
phubbing was associated with social disconnection. The first model investigated whether the
relationship between phubbing and perceived disconnection was source-specific by running a SEM
in which only the first-order factors of both the constructs were included. Specifically, in addition to
the measurement models of phubbing and social disconnection, SD-M and SD-F were regressed on
both PPS-M and PPS-F. The model yielded good fit indices [χ²(344) = 1909.25, p < .001; CFI =
.962; TLI = .958; RMSEA = .037; SRMR = .031]. As displayed in Figure 2, all the regression
coefficients were positive and significant at p < .001, except for SD-M on PPS-F that had a
probability of p = .014. Although phubbing from both mother and father was associated with the
perception of disconnection from both parents, the source-specificity of these relationships was
confirmed by testing a series of model constraints. The influence of PPS-M on SD-M was
significantly stronger than that of PPS-F on SD-M (Δb = .15, p < .001), as well as the influence of
PPS-F on SD-F compared to the one of PPS-M on SD-F (Δb = .08, p < .001). Consistently, the
influence of PPS-M on SD-M was stronger than the influence of PPS-M on SD-F (Δb = .12, p <
.001) as well as the influence of PPS-F on SD-F compared to the one of PPS-F on SD-M (Δb = .11,
p < .001). The percentage of explained variance was 16% for SD-M and 15% for SD-F.
[Figure 2 near here]
The second model tested whether parental phubbing was associated to the overall parental
disconnection, along with source-specific (i.e., mother and father) associations between phubbing
Running head: the Parental Phubbing Scale
19
and social disconnection. Results, graphically depicted in Figure 3, indicated a good model fit
[χ²(345) = 1915.42, p < .001; CFI = .962; TLI = .958; RMSEA = .037; SRMR = .032]. The highest
standardized regression coefficient was observed for the association between the two second-order
factors, β = .29, p < .001, followed by the association between PPS-M and SD-M, β = .28, p < .001,
and PPS-F and SD-F, β = .26, p < .001. However, model constraints between all possible pairs of
regression coefficients yielded non-significant results (SD on PPS vs. SD-M on PPS-M: Δb = 0.01,
p = .80; SD on PPS vs. SD-F on PPS-F: Δb = 0.04, p = .33; SD-M on PPS-M vs. SD-F on PPS-F:
Δb = 0.03, p = .14), showing equal magnitude. The highest R2 was observed for SD-F (.65),
followed by SD-M (.51) and SD (.09).
[Figure 3 near here]
Group differences in perceived parental phubbing
Two multilevel distinct MIMIC models have been estimated: one for the two sub-dimension
of PPS-M and PPS-F (M1) and another one focused on the overall PPS latent construct (M2). Both
models specifications resulted in values of alternative fit indices indicating close fit to the data [M1:
χ²(236) = 1475.9, p < .001; CFI = .949; TLI = .943; RMSEA = .042; SRMR = .027. M2: χ²(212) =
1496.8, p < .001; CFI = .949; TLI = .943; RMSEA = .044; SRMR = .029]. As for regression
coefficients, we found that males perceived to be less phubbed by parents than females and, at the
same time, first and second generation of migrants declared to be more exposed to this phenomenon
than natives (Table 3). These results were observed for both PPS-M and PPS-F. Although at a first
glance the influence of gender and ethnic origin seemed to be stronger for PPS-M than for PPS-F, a
direct comparison of the regression parameters yielded non-significant results, demonstrating that
gender (Δb = 0.02, p = .61) and ethnic origin (Δb = -0.04, p = .41) had the same effect on perceived
phubbing from both parents. Conversely, no significant differences in perceived phubbing were
found for the education level of parents.
[Table 3 near here]
Running head: the Parental Phubbing Scale
20
The analysis conducted on the overall construct of PPS (M2) confirmed that girls and
migrant students perceived to be more phubbed by parents than males and natives, while parental
education level does not appear to be a relevant predictor of phubbing at the family level.
General Discussion
Human beings have a fundamental need to belong (Baumeister & Leary, 1995); therefore,
they are constantly motivated to seek social connections with other individuals to satisfy such need.
Digital technologies, smartphones, in particular, offer several attractive ways to fulfill this basic
need, providing easy channels to create and maintain connections with people irrespectively of
space and time. Indeed, international surveys show how widespread is the use of such technologies
worldwide. In 2019, 66.6% of the world population owned a smartphone (+2.0% from 2018) and
spent more than 3 hours a day (+4.3% from 2018) using it (We Are Social, 2019).
Within this context, the present study was conceived to investigate phubbing within the
parent-child relationship. If being phubbed leads to feelings of relational devaluation, the threat to
fundamental psychological needs, and even depression (Chotpitayasunondh & Douglas, 2018;
Hales et al., 2018; Roberts & David, 2016; Wang et al., 2017), these effects could be stronger and
potentially more detrimental in the long-term when adolescents are phubbed by their parents. This
is because communication and parental responsiveness have central roles in children and
adolescents’ development (Baumrind, 1991; Caughlin & Malis, 2004; Davidov & Grusec, 2006;
Kochanska & Aksan, 2004; Pinquart, 2016). However, the adverse consequences of parental
phubbing cannot be examined without the proper measurement of the phenomenon. That measure is
still lacking; thus, our study aimed at filling in this gap by developing a brief scale of parental
phubbing as a preliminary and necessary step in the investigation of this phenomenon.
The results showed that the Parental Phubbing Scale (PPS) is a psychometrically valid
measure of adolescents’ perception of parental phubbing. The PPS consisted of two highly reliable
Running head: the Parental Phubbing Scale
21
sub-dimensions of phubbing (i.e., phubbing of mother and father) that identify an overall dimension
of parental phubbing. The PPS demonstrated full measurement and structural invariance for a set of
adolescents’ and parents’ characteristics, indicating that perceived parental phubbing is reliably
measured irrespectively of adolescents’ gender, ethnic origins, and mother’s and father’s education
level.
The PPS also demonstrated a good concurrent validity. Generally speaking, the perception
of phubbing was significantly and positively associated with the feeling of social disconnection
from parents. This association held for both the overall measure of parental phubbing and its
source-specific components (i.e., phubbing of mother and father). Digging into these associations
yielded some insights about how parental phubbing is structured. Phubbing is a construct that
primarily emerges in one-to-one interactions; thus its consequences (i.e., an increase of social
disconnection) should be observed in a specific relationship between two persons (e.g., the
adolescent and one specific parent). This is consistent with our results that showed that the
association between phubbing and social disconnection was significantly higher when both the
constructs were referred to the same parent than when perceived phubbing was referred to one
parent and the feeling of disconnection to the other one. However, the two relationships considered
(i.e., adolescent-mother and adolescent-father) are elements of the same meaningful social context,
namely the family unit. Thus, the relational dynamics of the two dyads are likely to be influenced
by one another, and the dyadic nature of the constructs at stake might be extended to the larger
context of the family, making it possible to estimate the overall constructs of parental phubbing and
parental social disconnection. Nevertheless, along with the possibility to estimate global parental
dimensions, the two constructs primarily concern one-to-one relationships, and this clearly emerged
in our analyses. Indeed, the parent-specific associations between phubbing and social disconnection
came out as fundamental paths to properly describe the link between adolescents’ perception of
Running head: the Parental Phubbing Scale
22
being ignored by their parents because of the smartphone and their perceived social distance from
their mother and father.
Generally speaking, the positive association between parental phubbing and social
disconnection is in line with studies that shed light on phubbing as an instance of social exclusion
(Chotpitayasunondh & Douglas, 2018; Hales et al., 2018). However, the present study is the first
showing this link in the parent-child relationship. Although the literature on technoference has
paved the way to the investigation of the adverse effects of parents distracted with technological
devices (McDaniel & Radesky, 2018; Stockdale et al., 2018), no studies have focused yet on
whether phubbed children would feel socially excluded by their parents. Given the importance of
the quality of the parent-child relationship, knowing that children being phubbed by their parents
feel excluded by them is extremely important. Indeed, feeling disconnected from others is one of
the biggest problems in our society (de Jong Gierveld, Van Tilburg, & Dykstra, 2006) and it is
related to a wide array of adverse consequences for physical health, cognitive functioning, and
emotional sphere (for an overview, see Riva and Eck, 2016). Concerning the parent-child
relationship, it is well-known how parenting style affects infants and adolescents' development. For
instance, two recent meta-analyses conducted on more than 1,000 studies each (Pinquart, 2016,
2017) found out that low parental responsiveness and neglectful parenting style were associated
with children’s externalizing problems and worse academic performance. These dimensions of
parenting cover a broad spectrum of cognitive, affective, social, and behavioral characteristics.
Accordingly, we argue that parental phubbing might be one of the many indicators that identify low
responsive or neglectful practices.
Finally, we discuss the differences accounted for by sociodemographic variables. Female
and non-native adolescents were more sensitive to both parental and parent-specific phubbing, and
no differences emerged between being phubbed by mother or father. This result allows us to
generalize the effect of gender and ethnic origin, claiming that females and first and second
Running head: the Parental Phubbing Scale
23
generation of migrants are more susceptible to parental phubbing. However, these findings do not
mean that parents are more likely to phub their children if the latter were females or migrants, but
only that females and migrant adolescents perceive to be more phubbed by their parents. The design
of the present research does not allow us to look for the reasons behind these results. Concerning
family educational background, we did not find any significant effect on children’s perceived
phubbing. Although the literature on digital inequality has found solid associations between
education level and both perceived digital overuse (Gui & Büchi, 2018) and digital parental
mediation style (Livingstone et al., 2015; Nikken & Opree, 2018), in our sample it seems that
educational differences do not translate into a different perception of being phubbed.
The main limitation of the present study concerns its correlational nature. Future studies
should adopt other methodologies (e.g., experimental, longitudinal) to explore causal links between
parental phubbing on adolescents’ development and psychological health, to uncover the underlying
processes and limit or reduce the adverse effects of this practice. Future studies should also consider
younger children to investigate the effects of parental phubbing at different stages of development.
Finally, research is needed further to specify the association between parental phubbing and
sociodemographic variables.
Conclusions
What are the consequences of the current massive use of smartphones on people’s social
lives? The present study proposed a new measure that can account for adolescents’ perception of
experiencing phubbing from their parents, the Parental Phubbing Scale (PPS). The large sample
size, the rigorousness of the analyses, and the quality of the results obtained make the PPS a reliable
and valid instrument for the research on the underinvestigated phenomenon of parental phubbing.
Moreover, the positive association between parental phubbing and social disconnection from
parents confirmed the link between phubbing and social exclusion.
Running head: the Parental Phubbing Scale
24
Waytz and Gray (2018) have recently tried to theorize the conditions under which
technology can impact sociability. The authors claimed that technology might enhance sociability
when devices and social media are used to both complement pre-existing, deep offline relationships,
or to maintain them when face-to-face interactions are otherwise difficult to attain. Conversely,
technology might impair sociability when superficial online interactions supplant deeper face-to-
face relationships. However, phubbing, despite having clear negative effects on a person's sociality
as reviewed above (e.g., Roberts & David, 2016; Wang et al., 2017), does not necessarily fall into
this case. In fact, phubbing might also occur when a "superficial" offline interaction is disrupted by
someone communicating online (e.g., via text messages) with a significant other (e.g., the partner or
a family member). In this sense, inspired by the effects of phubbing, we argue that the Waytz and
Gray’s principle about the negative impact of technology on sociability should be rephrased in
broader terms as following: technology might impair sociability when offline interactions are
disrupted by online ones. In our opinion, the issue is not in the superficiality or depth of the
interactions that can occur online and offline, but in the disruption of ongoing offline interactions
due to incoming online ones. Greater efforts are needed to explore the impact of these emerging
technologies on the ways humans connect each other.
Running head: the Parental Phubbing Scale
25
References
Abeele, M. M. P. V, Antheunis, M. L., & Schouten, A. P. (2016). The effect of mobile messaging
during a conversation on impression formation and interaction quality. Computers in Human
Behavior, 62, 562–569. https://doi.org/10.1016/j.chb.2016.04.005
Al-Saggaf, Y., & MacCulloch, R. (2018). Phubbing: How frequent? Who is phubbed? In which
situation? And using which apps? Proceedings of the Thirty Ninth International Conference on
Information Systems (ICIS), San Fransisco (CA). Retrieved from
https://aisel.aisnet.org/icis2018/behavior/Presentations/25
Asparouhov, T. (2006). General multi-level modeling with sampling weights. Communications in
Statistics - Theory and Methods, 35(3), 439–460. https://doi.org/10.1080/03610920500476598
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments
as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529.
https://doi.org/0033-2909
Baumrind, D. (1991). The influence of parenting style on adolescent competence and substance use.
Journal of Early Adolescence, 11(1), 56–95. https://doi.org/10.1177/0272431691111004
Bentler, P. M., & Bonnet, D. C. (1980). Significance tests and goodness of fit in the analysis of
covariance structures. Psychological Bulletin, 88(3), 588–606. https://doi.org/10.1037/0033-
2909.88.3.588
Bernstein, M. J. (2016). Research in social psychology: Consequences of short- and long-term
social exclusion. In P. Riva & J. Eck (Eds.), Social exclusion: Psychological approaches to
understanding and reducing its impact (pp. 51–72). New York, NY: Springer.
Brown, T. A. (2015). Confirmatory factor analysis for applied research. New York, NY: Guilford
Press.
Running head: the Parental Phubbing Scale
26
Byrne, B. M. (1988). The Self Description Questionnaire III: testing for equivalent factorial validity
across ability. Educational and Psychological Measurement, 48(2), 397–406.
https://doi.org/10.1177/0013164488482012
Caughlin, J. P., & Malis, R. S. (2004). Demand/withdraw communication between parents and
adolescents: Connections with self-esteem and substance use. Journal of Social and Personal
Relationships, 21(1), 125–148. https://doi.org/10.1177/0265407504039843
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing
measurement invariance. Structural Equation Modeling, 9(2), 233–255.
https://doi.org/10.1207/S15328007SEM0902_5
Chotpitayasunondh, V., & Douglas, K. M. (2016). How “phubbing” becomes the norm: The
antecedents and consequences of snubbing via smartphone. Computers in Human Behavior,
63, 9–18. https://doi.org/10.1016/j.chb.2016.05.018
Chotpitayasunondh, V., & Douglas, K. M. (2018). The effects of “phubbing” on social interaction.
Journal of Applied Social Psychology, 48, 304–316. https://doi.org/10.1111/jasp.12506
Cialdini, R. B. (1993). Influence: The psychology of persuasion. New York, NY: Morrow.
Davidov, M., & Grusec, J. E. (2006). Untangling the links of parental responsiveness to distress and
warmth to child outcomes. Child Development2, 77(1), 44–58. https://doi.org/10.1111/j.1467-
8624.2006.00855.x
de Jong Gierveld, J., Van Tilburg, T. G., & Dykstra, P. A. (2006). Loneliness and social isolation.
In D. Perlman & A. Vangelisti (Eds.), The Cambridge Handbook of Personal Relationships.
Cambridge, UK: Cambridge University Press.
Falk, A., & Fischbacher, U. (2006). A theory of reciprocity. Games and Economic Behavior, 54(2),
293–315. https://doi.org/10.1016/j.geb.2005.03.001
Running head: the Parental Phubbing Scale
27
Gentile, D. A., Coyne, S. M., & Bricolo, F. (2013). Pathological technology addictions: What is
scientifically known and what remains to be learned. In K. E. Dill (Ed.), The Oxford handbook
of media psychology (pp. 382–402). New York, NY: Oxford University Press.
https://doi.org/10.1093/oxfordhb/9780195398809.013.0022
Gui, M., & Büchi, M. (2019). From use to overuse: Digital inequality in the age of communication
abundance. Social Science Computer Review. Advanced online publication.
https://doi.org/10.1177/0894439319851163
Gui, M., Gerosa, T., Garavaglia, A., Petti, L., & Fasoli, M. (2019). Digital well-being – Schools.
validation of a digital media education programme in high schools. Retrieved from
https://www.benesseredigitale.eu/materiali
Hales, A. H., Dvir, M., Wesselmann, E. D., Kruger, D. J., & Finkenauer, C. (2018). Cell phone-
induced ostracism threatens fundamental needs. The Journal of Social Psychology, 158(4),
460–473. https://doi.org/10.1080/00224545.2018.1439877
Joreskog, K. G., & Goldberger, A. S. (1975). Estimation of a model with multiple indicators and
multiple causes of a single latent variable. Journal of the American Statistical Association,
70(351), 631–639. https://doi.org/10.2307/2285946
Karadağ, E., Tosuntaş, S. B., Erzen, E., Duru, P., Bostan, N., Şahin, B. M., … Babadağ, B. (2015).
Determinants of phubbing, which is the sum of many virtua addictions: A structural equation
model. Journal of Behavioral Addictions, 4(2), 60–74.
https://doi.org/10.1556/2006.4.2015.005
Kardefelt-Winther, D., Heeren, A., Schimmenti, A., van Rooij, A., Maurage, P., Carras, M., …
Billieux, J. (2017). How can we conceptualize behavioural addiction without pathologizing
common behaviours? Addiction, 112(10), 1709–1715. https://doi.org/10.1111/add.13763
Kim, E. S., Kwok, O.-M., & Yoon, M. (2012). Testing factorial invariance in multilevel data: A
Running head: the Parental Phubbing Scale
28
Monte Carlo study. Structural Equation Modeling: A Multidisciplinary Journal2, 19(2), 250–
267. https://doi.org/10.1080/10705511.2012.659623
King, R. C., & Dong, S. (2017). The impact of smartphone on young adults. The Business and
Management Review, 8, 342–349. Retrieved from
http://www.abrmr.com/myfile/best_track/conference_44279.pdf
Kline, R. B. (2015). Principles and practice of structural equation modeling. New York, NY:
Guilford Publications.
Kochanska, G., & Aksan, N. (2004). Development of mutual responsiveness between parents and
their young children. Child Development, 75(6), 1657–1676. https://doi.org/10.1111/j.1467-
8624.2004.00808.x
Kwon, M., Kim, D.-J., Cho, H., & Yang, S. (2013). The smartphone addiction scale: development
and validation of a short version for adolescents. PloS One, 8(12), e83558.
https://doi.org/10.1371/journal.pone.0083558
Leary, M. R. (2001). Toward a conceptualization of interpersonal rejection. In M. R. Leary (Ed.),
Interpersonal rejection (pp. 3–20). Oxford, UK: Oxford University Press.
Leary, M. R. (2020). The need to belong, the sociometer, and the pursuit of relational value:
Unfinished business. Self and Identity. https://doi.org/10.1080/15298868.2020.1779120
Lee, Y.-K., Chang, C.-T., Lin, Y., & Cheng, Z.-H. (2014). The dark side of smartphone usage:
Psychological traits, compulsive behavior and technostress. Computers in Human Behavior,
31, 373–383. https://doi.org/10.1016/j.chb.2013.10.047
Lin, Y. H., Chiang, C. L., Lin, P. H., Chang, L. R., Ko, C. H., Lee, Y. H., & Lin, S. H. (2016).
Proposed diagnostic criteria for smartphone addiction. PLoS ONE, 11(11), e0163010.
https://doi.org/10.1371/journal.pone.0163010
Running head: the Parental Phubbing Scale
29
Livingstone, S., Mascheroni, G., Dreier, M., Chaudron, S., & Lagae, K. (2015). How parents of
young children manage digital devices at home: The role of income, education and parental
style. London: EU Kids Online, LSE. Retrieved from: http://eprints.lse.ac.uk/63378/
Macquarie. (2013). Phubbing. In Macquarie Dictionary (6th ed.). MacQuarie Library.
McDaniel, B. T., & Coyne, S. M. (2016). “Technoference”: The interference of technology in
couple relationships and implications for women’s personal and relational well-being.
Psychology of Popular Media Culture, 5(1), 85–98. https://doi.org/10.1037/ppm0000065
McDaniel, B. T., & Radesky, J. S. (2018). Technoference: Parent distraction with technology and
associations with child behavior problems. Child Development, 89(1), 100–109.
https://doi.org/10.1111/cdev.12822
Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance.
Psychometrika, 58(4), 525–543. https://doi.org/10.1007/BF02294825
Muthén, M., & Muthén, B. O. (2015). Mplus user’s guide. Seventh edition. Los Angeles, CA:
Muthén & Muthén.
Nikken, P., & Opree, S. J. (2018). Guiding young children’s digital media use: SES-differences in
mediation concerns and competence. Journal of Child and Family Studies, 27(6), 1844–1857.
https://doi.org/0.1007/s10826-018-1018-3
Nuñez, T. R., Grewe, A., Trienens, L.-M., Kowalczyk, C., Nitschke, D., Tran, B. T. F., & Eimler, S.
C. (2018). Watching phubbers hurts: Observing phone-snubbing in others’ social interactions
affects observers’ well-being. Poster session presented at the 51st Congress of the German
Association for Psychology (DGPs), Frankfurt, D. Retrieved from
https://www.researchgate.net/publication/327722078_Watching_Phubbers_Hurts_Observing_
Phone-Snubbing_in_Others%27_Social_Interactions_Affects_Observers%27_Wellbeing
Running head: the Parental Phubbing Scale
30
Pancani, L., Preti, E., & Riva, P. (2020). The psychology of smartphone: The development of the
Smartphone Impact Scale (SIS). Assessment, 27(6), 1176-1197.
https://doi.org/10.1177/1073191119831788
Pinquart, M. (2016). Associations of parenting styles and dimensions with academic achievement in
children and adolescents: A meta-analysis. Educational Psychology Review, 28(3), 475–493.
https://doi.org/10.1007/s10648-015-9338-y
Pinquart, M. (2017). Associations of parenting dimensions and styles with externalizing problems
of children and adolescents: An updated meta-analysis. Developmental Psychology, 53(5),
873–932. https://doi.org/10.1037/dev0000295
Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance conventions and reporting: The
state of the art and future directions for psychological research. Developmental Review, 41,
71–90. https://doi.org/10.1016/j.dr.2016.06.004
Riva, P., & Eck, J. (2016). Social exclusion: Psychological approaches to understanding and
reducing its impact. New York, NY: Springer.
Riva, P., Montali, L., Wirth, J. H., Curioni, S., & Williams, K. D. (2017). Chronic social exclusion
and evidence for the resignation stage: An empirical investigation. Journal of Social and
Personal Relationships, 34(4), 541–564. https://doi.org/10.1177/0265407516644348
Roberts, J. A., & David, M. E. (2016). My life has become a major distraction from my cell phone:
Partner phubbing and relationship satisfaction among romantic partners. Computers in Human
Behavior, 54, 134–141. https://doi.org/10.1016/j.chb.2015.07.058
Roberts, J. A., Williams, B. H., & David, M. E. (2017). Put down your phone and listen to me: How
boss phubbing undermines the psychological conditions necessary for employee engagement.
Computers in Human Behavior, 75, 206–217. https://doi.org/10.1016/j.chb.2017.05.021
Running head: the Parental Phubbing Scale
31
Stockdale, L. A., Coyne, S. M., & Padilla-Walker, L. M. (2018). Parent and child technoference and
socioemotional behavioral outcomes: A nationally representative study of 10- to 20-year-old
adolescents. Computers in Human Behavior, 88, 219–226.
https://doi.org/10.1016/j.chb.2018.06.034
Van Dijk, J. A. (2005). The deepening divide: Inequality in the information society. Thousand
Oaks, CA: Sage.
Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance
literature: Suggestions, practices, and recommendations for organizational research.
Organizational Research Methods, 3(1), 4–70. https://doi.org/10.1177/109442810031002
Wang, J., & Wang, X. (2012). Structural equation modeling: Applications using Mplus. Chichester,
UK: John Wiley & Sons.
Wang, X., Cie, X., Wang, Y., Wang, P., & Lei, L. (2017). Partner phubbing and depression among
married Chinese adults: The roles of relationship satisfaction and relationship length.
Personality and Individual Differences, 110, 12–17. https://doi.org/10.1016/j.paid.2017.01.014
Waytz, A., & Gray, K. (2018). Does online technology make us more or less sociable? A
preliminary review and call for research. Perspectives on Psychological Science, 13(4), 473-
491. https://doi.org/10.1177/1745691617746509
We Are Social. (2019). Global digital report 2019. Retrieved from https://wearesocial.com/global-
digital-report-2019
Widaman, K. F., & Reise, S. P. (1997). Exploring the measurement invariance of psychological
instruments: applications in the substance use domain. In K. J. Bryant, M. E. Windle, & S. G.
West (Eds.), The science of prevention: methodological advances from alcohol and substance
abuse research (pp. 281–324). Whashington, DC.
Running head: the Parental Phubbing Scale
32
Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452.
https://doi.org/10.1146/annurev.psych.58.110405.085641
Williams, K. D. (2009). Ostracism: A temporal need-threat model. In M. P. Zanna (Ed.), Advances
in experimental social psychology (Vol. 41, pp. 275–314). San Diego, CA: Elsevier Academic
Press.
Wu, J.-Y., & Kwok, O.-M. (2012). Using SEM to analyze complex survey data: A comparison
between design-based single-level and model-based multilevel approaches. Structural
Equation Modeling: A Multidisciplinary Journal, 19(1), 16–35.
https://doi.org/10.1080/10705511.2012.634703
Running head: the Parental Phubbing Scale
33
Table 1. Sample characteristics (N = 3,289): Descriptive statistics of students’ sociodemographic
and school characteristics.
Variable (value)
Mean (SD)
Frequency (%)
Age
15.2
(0.6)
Missing
1
(0.0)
Gender
Male (1)
1,586
(48.2)
Female (2)
1,699
(51.7)
Missing
4
(0.1)
Ethnic origins
Native (1)
2,859
(87.2)
Other country (2)
420
(12.8)
Missing
10
(0.3)
Mother educational level
Low (1)
748
(22.7)
Middle (2)
1,548
(47.1)
High (3)
854
(26.0)
Missing
96
(2.9)
Father educational level
Low (1)
1,000
(30.4)
Middle (2)
1,345
(40.9)
High (3)
769
(23.4)
Missing
175
(5.3)
Type of school
Lyceum (1)
1,739
(52.9)
Technical Institute (2)
1,106
(33.6)
Professional Institute (3)
444
(13.5)
Missing
0
(0.0)
Running head: the Parental Phubbing Scale
34
Table 2. The results of the measurement invariance analyses by children’s gender and ethnic
origins, and parents’ education levels.
Fit Indices (MLMV estimation method)
χ2
df
χ2
p-val
RMSEA
[90% C.I.]
CFI
TLI
Mc
Δχ2
p-val
ΔCFI
ΔMc
Gender
Baseline models
males
470.7
69
<.001
.062 [.056-.067]
.951
.935
–
–
–
–
females
429.7
69
<.001
.056 [.051-.061]
.961
.949
–
–
–
–
Configural
900.5
138
<.001
.059 [.055-.063]
.956
.942
.655
–
–
–
Weak
923.4
150
<.001
.057 [.053-.060]
.956
.946
.658
.104
.000
-.003
Strong
1039.9
164
<.001
.058 [.055-.061]
.950
.944
.656
<.001
.006
.002
Residual cov.
1042.6
171
<.001
.057 [.053-.060]
.950
.947
.658
0.097
.000
-.002
Structural
1070.8
174
<.001
.057 [.054-.060]
.949
.9496
.658
<.001
.001
.000
Ethnic origins
Baseline models
natives
713.4
69
<.001
.058 [.054-.062]
.957
.943
–
–
–
–
others
178.6
69
<.001
.063 [.052-.074]
.952
.937
–
–
–
–
Configural
898.0
138
<.001
.059 [.055-.062]
.956
.942
.655
–
–
–
Weak
931.7
150
<.001
.057 [.054-.061]
.955
.945
.657
.056
.001
.002
Strong
990.7
164
<.001
.056 [.053-.060]
.952
.947
.659
<.001
.003
.002
Residual cov.
990.1
171
<.001
.055 [.052-.058]
.953
.950
.661
.500
-.001
.002
Structural
998.5
174
<.001
.054 [.051-.058]
.952
.950
.662
.027
.001
.001
Mother’s education level
Baseline models
low
271.6
69
<.001
.064 [.056-.072]
.953
.938
–
–
–
–
middle
423.3
69
<.001
.058 [.053-.064]
.956
.943
–
–
–
–
high
243.6
69
<.001
.055 [.048-.063]
.959
.946
–
–
–
–
Configural
932.4
207
<.001
.059 [.055-.062]
.956
.942
.677
–
–
–
Weak
969.9
231
<.001
.056 [.052-.060]
.955
.947
.683
.478
.001
.006
Strong
1015.5
259
<.001
.053 [.050-.057]
.954
.952
.689
.747
.001
.006
Residual cov.
1023.7
273
<.001
.052 [.049-.055]
.955
.955
.693
.425
-.001
.004
Structural
1030.5
279
<.001
.051 [.048-.055]
.955
.956
.694
.134
.000
.001
Father’s education level
Baseline models
low
318.0
69
<.001
.061 [.054-.068]
.954
.939
–
–
–
–
middle
378.8
69
<.001
.058 [.053-.064]
.957
.944
–
–
–
–
high
234.5
69
<.001
.057 [.049-.065]
.958
.945
–
–
–
–
Configural
929.3
207
<.001
.059 [.055-.062]
.956
.943
.678
–
–
–
Weak
975.8
231
<.001
.056 [.053-.060]
.955
.947
.683
.053
.001
.005
Strong
1040.7
259
<.001
.055 [.051-.058]
.953
.950
.687
.011
.002
.004
Residual cov.
1058.2
273
<.001
.053 [.050-.057]
.953
.953
.690
.067
.000
.003
Structural
1076.5
279
<.001
.053 [.050-.056]
.952
.953
.690
.004
.001
.000
Running head: the Parental Phubbing Scale
35
Table 3. The results of the MIMIC models. Standardized estimates (and standard errors) are
reported.
Model 1
Model 2
PPS-M
PPS-F
PPS
Student characteristics
Gender (ref. female)
males
-0.071
(0.021)**
-0.058
(0.022)**
-0.109
(0.033)**
Ethnic origins (ref. natives)
others
0.074
(0.020)**
0.053
(0.020)**
0.103
(0.028)**
Parents characteristics
Mother’s education (ref. low)
middle
-0.021
(0.025)
-
-
-
-
high
-0.040
(0.023)
-
-
-
-
Father’s education (ref. low)
middle
-
-
-0.018
(0.022)
-
-
high
-
-
-0.002
(0.021)
-
-
Parents’ education (ref. low)
middle
-
-
-
-
-0.022
(0.043)
high
-
-
-
-
-0.039
(0.044)
p-value: * ≤ 0.05 ** ≤ 0.01
Running head: the Parental Phubbing Scale
36
Figure 1 – The results of the confirmatory factor analysis on the PPS: Standardized parameters are
displayed.
Note: PPS = overall dimension of parental phubbing, PPS-M = phubbing perceived from mother,
PPS-F = phubbing perceived from father. All the parameters were significant at level p < .05.
Running head: the Parental Phubbing Scale
37
Figure 2 – The results of the structural equation model testing the association between source-
specific dimensions of parental phubbing and perceived disconnection from parents: Standardized
parameters are displayed.
Note: PPS-M = phubbing perceived from mother, PPS-F = phubbing perceived from father, SD-M
= social disconnection perceived from mother; SD-F = social disconnection perceived from father.
All the parameters were significant at level p < .05.
Running head: the Parental Phubbing Scale
38
Figure 3 – The results of the structural equation model testing the association between overall
dimensions of parental phubbing and perceived disconnection from parents: Standardized
parameters are displayed.
Note: PPS = overall dimension of parental phubbing, PPS-M = phubbing perceived from mother,
PPS-F = phubbing perceived from father, SD = overall dimension of social disconnection from
parents, SD-M = social disconnection perceived from mother; SD-F = social disconnection
perceived from father. All the parameters were significant at level p < .05.
Running head: the Parental Phubbing Scale
39
Supplementary Material
Factor Validity
Factor validity concerns the extent to which the selected observable items adequately cover
the model specification of the latent construct(s) being studied. The model specification is a set of
equations that should reproduce the theoretical relationships across variables; in the context of factor
validity, these relationships mainly include loadings of observed variables on latent ones (i.e.,
constructs) and correlations among the latters. In other words, testing for the factor validity of a
theoretical construct concerns defining the model specifications that best fits the available data. The
statistical technique of Confirmatory Factor Analysis (CFA) is the most appropriate to test factor
validity and compare different model specifications, thus it has been chosen to conduct these analyses.
Measurement and Structural Invariance
The measurement invariance test allowed to check whether the psychometric properties of the
latent construct and, therefore, the equations used to create the latent factor scores can be considered
equal across sub-populations of interest (1). One of the major threats to this assumption is represented
by the risk of measurement bias. It consists in a potential difference between the estimated and the
true parameter resulting from the presence of a nuisance factor that produces an undesirable source
of measurement variance (2,3). If not seriously taken into account, this kind of bias could drive to
inaccurate inferences about the results of any comparative analysis, especially if it works differently
on the different sub-populations of interest (1). Considering that one of the purposes of this study was
to properly quantify the average differences on PPS-M and PPS-F across groups, we kept under
control these interfering factors applying the standard procedure for testing measurement invariance
through the Multi-Group Confirmatory Factor Analysis (MG-CFA) (4,5). The same technique has
also been adopted to evaluate the degree of variability in the first order latent factors variance (i.e.,
PPS-M and PPS-F) and their correlational relationships across groups of students. This additional
Running head: the Parental Phubbing Scale
40
analysis, commonly referred as structural invariance test in the literature (6,7), allowed to check the
cross-group stability in both the distribution of PPM-F and PPM-M and the way they relate to each
other. Both measurement and structural invariance tests are based on the comparison of the fit of a
series of hierarchically nested models. In each step of the analysis, an increasing number of equality
constraints were fixed among the estimated parameters, allowing to check the presence of any
significant difference in the model specification across groups.
References
1. Sass D. Testing measurement invariance and comparing latent factor means within a
confirmatory factor analysis framework. J Psychoeduc Assess. 2011;29(4):347–63.
2. Van de Vijver FJR, Poortinga YH. Conceptual and methodological issues in adapting tests.
In: Hambleton RK, Merenda PF, Spielberger CD, editors. Adapting educational and
psychological tests for cross-cultural assessment. Mahwah, NJ; 2005. p. 39–63.
3. Gerosa T. Measuring adolescents’ affective civic competence: Validity and cross-group
equivalence of a second-order hierarchical latent construct. Appl Res Qual Life.
2019;14(2):335–58.
4. Vandenberg RJ, Lance CE. A review and synthesis of the measurement invariance literature:
Suggestions, practices, and recommendations for organizational research. Organ Res
Methods. 2000;3(1):4–70.
5. Chen FF, Sousa KH, West SG. Teacher’s corner: testing measurement invariance of second-
order factor models. Struct Equ Model. 2005;12(3):471–92.
6. Putnick DL, Bornstein MH. Measurement invariance conventions and reporting: The state of
the art and future directions for psychological research. Dev Rev. 2016;41:71–90.
7. Dimitrov DM. Testing for factorial invariance in the context of construct validation. Meas
Running head: the Parental Phubbing Scale
41
Eval Couns Dev. 2010;43(2):121–49.
