Psychometric modeling of the pervasive use of Facebook through
psychophysiological measures: Stress or optimal experience?
Pietro Cipresso
a,⇑, Silvia Serino
a, Andrea Gaggioli
a,b, Giovanni Albani
c, Alessandro Mauro
c,
Giuseppe Riva
a,ba
Applied Technology for Neuro-Psychology Lab, IRCCS Istituto Auxologico Italiano, Via Magnasco 2, 20149 Milano, Italy
bPsychology Department, Catholic University of Milan, Largo Gemelli, 1, 20123 Milan, Italy c
Division of Neurology and Neurorehabilitation, San Giuseppe Hospital, IRCCS Istituto Auxologico Italiano, Piancavallo, VB, Italy
a r t i c l e
i n f o
Article history:
Keywords:
Psychometric modeling Psychophysiology Social network sites Mobile social networking Facebook
Psychological stress
a b s t r a c t
Social network sites have been studied extensively with an aim to understand users’ experience during their use. Facebook is the most frequently used social network site, with more than one billion active users. However, despite 819 million of active Facebook users currently accessing it using their mobile device, only a limited number of studies have investigated their experiences with Facebook. Our goal was to objectively verify in an experimental setting the subjective experience of users accessing Facebook through a PC and through a smart phone, which has a higher pervasiveness. Psychophysiological correlates of 28 subjects were measured using wearable biosensors, which record signals through an electrocardiogram; a 14-channel Electroencephalogram; facial Electromyography; an Electrooculogram, and a chest Respiration strip. An accurate signal processing permitted to compute twenty psychophysiological measures for the statistical analysis. The results showed significant patterns in arousal, valence, attention, and anxiety, indicating a subjects’ engagement during Facebook navigation, which was also more evident during the mobile session, suggesting an optimal experience during the per-vasive use of Facebook. Some hypotheses and the directions for future studies are presented, in particular, the suggestion to make further studies with higher ecological validity, outside the lab, to assess the ubi-quitous use of Facebook.
Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction
Social networking sites (SNSs), such as Facebook, Linkedin, Google+, and Twitter, are becoming increasingly prevalent in many aspects of communication, interaction, human behavior, and even personality (Cheung, Chiu, & Lee, 2011; Orr, Ross, Simmering, Arseneault, & Orr, 2009; Riva, Wiederhold, & Cipresso, 2015a, 2015b; Ross et al., 2009).
SNSs increase individuals’ visibility and availability within their social network (Boyd & Ellison, 2007; Riva, 2010). According to
Boyd and Ellison (2007), social networking sites could be defined as web-based services that enable users (1) to create a personal profile with different privacy level, (2) to structure a list of other users with whom they share a connection, and (3) to view and ana-lyze their behavior within the system and their list of connections.
People navigate through SNSs anywhere a computer is available to connect, including their homes, workplaces, or in public places. Furthermore, the recent development and diffusion of mobile plat-forms enhance the possibility to connect to the Internet and navi-gate through social networks (Ran & Lo, 2006). In particular, most contemporary SNSs have developed pages to navigate through mobile devices, and many of them have developed specific
applica-tions to improve mobile experiences (for example, http://
www.facebook.com/mobile/ for Facebook mobile App, http:// www.linkedin.com/static?key=mobile for LinkedIn, and http:// www.google.com/mobile/+/, for Google+). The most used SNS, as of June 30, 2013, is Facebook, with more than one billion active users (1.15 billion monthly active users at the end of June 2013); furthermore, there are about 819 million monthly active users (June 2013 data) currently accessing Facebook through their mobile devices (http://newsroom.fb.com/company-info/).
Navigating through SNSs using the PC and a mobile platform has two main differences: (1) mobile platform can be multifunc-tional, taking advantage of contextual affordances more than a
http://dx.doi.org/10.1016/j.chb.2015.03.068 0747-5632/Ó 2015 Elsevier Ltd. All rights reserved.
⇑Corresponding author.
E-mail address:p.cipresso@auxologico.it(P. Cipresso). URL:https://www.pietrocipresso.com(P. Cipresso).
Contents lists available atScienceDirect
Computers in Human Behavior
PC (Wood, 2011), and (2) mobile platform is – by definition – ubi-quitous (Zhou & Lu, 2011). Thus, mobile scenarios enable functions that cannot be used with a standard personal computer. In particu-lar, one mobile device, the mobile phone, has played a fundamental role in everyday life. Mobile phones have quickly evolved from only voice and text-based devices, which enable minimal user-de-vice interaction, to personal digital assistant, with digital camera, GPS, navigator, MP3 and video player, interactive agenda, clock and alarms, instant messaging, and Internet browser (Barkhuus & Polichar, 2011).
In this study, we were interested in examining subjects’ experi-ence during SNSs use, thus, we focused on three main aspects of time–space continuum of subjects’ states: physiological arousal, emotional valence, and sustained attention.
According to the classic valence-arousal model (Lang, 1995; Russell, 1979), to identify affective states of subjects during an experimental session, we can consider the two dimensions of ‘‘ac-tivation’’, namely, physiological arousal and emotional valence. Valence and arousal dynamics have been used extensively in psy-chophysiological research as an objective way to measure affective states during a mediated experience (Backs, da Silva, & Han, 2005; Bradley & Lang, 2000; Cipresso et al., 2012; Codispoti, Surcinelli, & Baldaro, 2008; Gruhn & Scheibe, 2008; Keil et al., 2002; Morris, 1995; Rubin, Rubin, Graham, Perse, & Seibold, 2009; Salimpoor, Benovoy, Larcher, Dagher, & Zatorre, 2011; Salimpoor, Benovoy, Longo, Cooperstock, & Zatorre, 2009; Tajadura-Jimenez, Pantelidou, Rebacz, Vastfjall, & Tsakiris, 2011). More recently, an extensive research has been done to discern different emotions by the means of cardiovascular measures (Cacioppo, Tassinary, & Berntson, 2007; Rainville, Bechara, Naqvi, & Damasio, 2006), con-firming the results that can be obtained using specific pattern of
the cardiovascular indexes (Magagnin et al., 2010; Mauri,
Cipresso, Balgera, Villamira, & Riva, 2011; Mauri et al., 2010). The third dimension used to assess the Facebook experience was the attention. According toDraper, Kaber, and Usher (1998), the presence occurs when more attentional resources are allocated to the computer-mediated environment.
If we are able to demonstrate that users are engaged during an experience and that during the same experience, their level of attention (Schupp et al., 2007) to the resource is high, then we can affirm that their sense of presence is high and this increases an optimal experience of subjects through an higher involvement (Riley, Kaber, & Draper, 2004; Riva, Davide, & Ijsselsteijn, 2003).
Thus, we assessed the experience by studying the physiological arousal, emotional valence, and sustained attention and by identifying, like in the classic valence-arousal model, the affective states to assess an experience effect due to a higher level of atten-tion (Fig. 1).
1.1. The psychophysiological assessment
As before stated several psychophysiological studies have widely extended the accuracy of the interpretation of the measures considered. Briefly, physiological arousal can be measured using Beta Waves observed using electroencephalogram (EEG) and gal-vanic skin response (GSR), using heart rate extracted from cardio-vascular activity through Electrocardiogram (ECG) or Blood Volume Pulse (BVP), and using respiration signal (RSP). Emotional valence can be measured using Alpha wave asymmetry on frontal channels of electroencephalogram (EEG), self-reports, facial expression identification, eye-blink startle, and facial EMG corrugator and/or facial EMG zygomatic. It has been broadly demonstrated that the facial EMG corrugator is one of the best
measures of emotional valence (Blumenthal et al., 2005).
Attention can be measured using Slow Alpha waves extracted from electroencephalogram (EEG), self-reports, Eye Tracking, Posture, and Behavioral analysis.
1.2. Research questions
Our study aim at investigating the pervasive use of Facebook by using psychophysiological measures. In particular, the following two research questions (RQ) emerges:
RQ1: Can Facebook navigation leads to an engagement state, aside from platform used and the related levels of pervasiveness? To answer this question, we used standard psychophysiological measures. We need to consider that the more the subjects will be engaged and attentive, the more likely they will be to achieve an optimal state characterized by positive valence, high arousal, and high attention. Thus, we were interested in objectively identify specific pattern of users’ affective state in the Arousal-Valence-Attention space while experiencing Facebook navigation trough a PC and a mobile platform.
RQ2: Can Facebook navigation through the mobile be more effective, leading to a higher pervasiveness able to induce an opti-mum state?
According to Favela and Colleagues (Favela, Tentori, & Gonzalez, 2010) two aspects should be evaluated in the experimental design of ubiquitous technologies, ecological validity and pervasiveness. Our study design considered the results of their research, specifi-cally the results regarding the pervasiveness of the navigation using a PC or a mobile. In fact, our study is related to a pervasive dimension between classic laboratory experiment and in vivo experiment at the highest level of pervasiveness. Favela and col-leagues identified this kind of experiment as an ex vivo study. Actually, the use of mobile, even if in a laboratory setting, gives a number of tools and capabilities that make the social networking experience more pervasive compared to a classic PC use (Coyne, 2010; Wood, 2011). Thus, our study aimed to compare a PC session, as a low pervasive one, with a Mobile session, that has a higher level of pervasiveness for the subject.
Fig. 1. Experience representation in the tridimensional space defined by physio-logical arousal, emotional valence, and attention.
Considerations for future study are given following, in the last part of the paper.
2. Materials and methods 2.1. Subjects
The participants were 28 subjects (16 females and 12 males) ranging in age from 19 to 55, with an average of 29. They were all volunteers from San Giuseppe Hospital in Piancavallo (VB), Italy recruited from the hospital staff (e.g., nurses, guardians, doc-tors). All subjects were experienced in the use of PC and smart phone. Exclusion criteria were related to the states of participants’ cardiac, optic, mental, and psychological health. Participants were asked not to drink caffeine or alcohol and not to smoke prior to the experimental test to avoid any effects of these substances on the central and autonomic nervous system.
All participant data were saved in a password protected file according to the criteria to protect personal health information (El Emam, Moreau, & Jonker, 2011) and encrypted with PsychoPass method (Cipresso, Gaggioli, Serino, Cipresso, & Riva, 2012).
2.2. Ethics statement
The scientific review board of the Applied Technology at the Neuro-Psychology IRCCS Istituto Auxologico Italiano has approved the study. All participants gave written informed consent to the experimental procedure according to the rules of the scientific review board.
2.3. Procedure
Subjects who met the experimental criteria were contacted face-to-face, via email, and/or telephone to schedule a meeting at the Neurophysiology Laboratory of the Department of Neurology located at San Giuseppe Hospital in Piancavallo (VB), Italy. They were welcomed by a specialized neurophysiologist technician and by a physician who assisted them for the duration of all lab-oratory sessions. The experimenters were instructed to maintain a neutral vocal tone and a neutral behavior while the subjects were being exposed to experimental stimuli. When subjects arrived at the Neurophysiology Laboratory, they were asked to sit down in front of a computer, and they were told about the general goals of the research investigation, the procedures to be used, and the information about the study objects. All subjects were required to sign a release form. While explaining the broad function of elec-trodes, the subjects were being prepared to wear the biosensors for the assessment of psychophysiological indexes. At this point, sub-jects undergone two counterbalanced sessions (PC, Mobile), each one with three randomized conditions (Relax, Facebook, and Stress).
In the PC session, the subjects seated in front of a computer monitor were exposed to the three PC conditions. In particular, in Relax condition subjects were asked to relax for five minute in front of a monitor showing panorama slide shows with a soft music in order to help them complete the relaxing exercise. In Facebook condition, subjects were asked to log into their Facebook account and freely navigate for five minute using the PC in front of them. In Stress condition, subjects were asked to execute a Stroop task, a cognitive task widely used to provide a ground truth for stress, which took five minutes to complete. In the Mobile session, the experimenters gave a smart phone (Apple iPhone 3Gs) to the sub-jects to complete the three Mobile conditions. In the Relax Mobile condition, subjects were asked to relax for five minutes with the
help of an app (iPhone application) that offered relaxing scenarios with soft and relaxing music. In the Facebook Mobile condition, subjects were asked to browse through their Facebook for five min-utes using the iPhone app developed by Facebook for this specific device. In the Stress Mobile condition, subjects were asked to engage in another cognitive task developed for the mobile device for five minute. Between the two sessions, the subjects were asked to keep their eyes closed for about ten minutes.
To summarize, the procedure consisted of two counterbalanced parts, one executed using a PC and the other executed using a mobile device, with a break in the middle in order to recover from the preceding session. Stimuli and tasks used in relax and stress conditions of both sessions were validated in many studies for eli-citing that states (Cipresso et al., 2012; Magagnin et al., 2010; Mauri et al., 2010, 2011).
At the end of the experimental session, the experimenter helped subjects remove all electrodes and patches while explaining the scientific rationale for using the stimuli and the aims of the experi-ment. Every recording was marked using a synchronization algo-rithm developed (by using Matlab program language) to align the stimulus with all psychophysiological signals (Cipresso, 2011). To improve the precision of such algorithm, subjects were asked to blink their eyes rapidly five times before each stimulus. Using elec-trodes near the eyes (Electrooculogram [EOG]) connected to all other biosensors, this operation guaranteed a precision of 1/100 of a second, which allowed for a better analysis and a more suc-cessful experiment. Once extracted, all biosignals were analyzed in MATLAB (The Mathworks, Inc.; Natick, MA) and branched in six conditions (categories): Relax, Facebook, Stress, Relax Mobile, Facebook Mobile, and Stress Mobile. Each condition contains all psychophysiological signals that can be processed for time and spectral analysis and other signal processing procedures in order to extract a series of indexes for the statistical analysis with the aid of the statistical software SPSS, version 17 (Statistical Package for the Social Sciences – SPSS for Windows, Chicago, Illinois, USA) and STATA (Stata statistical analysis software, version 11, Stata Corp, College Station, Tex).
2.4. Physiological signals
The responses of the central and peripheral nervous system were measured using an ECG (electrocardiogram); two facial EMG (electromyography), namely EMG-CS (corrugator supercilii) and EMG-Z (zygomatic major); EOG (Electrooculogram); and RSP (chest respiration). Signals were acquired by means of a
profes-sional SOMNOscreen PSG device (SOMNOmedics GmbH,
Randersacker, Germany) certified for medical use. Furthermore, we used an Epoc, a neuro-signal acquisition and processing wire-less neuroheadset device for the acquisition of 14 EEG
(electroen-cephalogram) channels (plus CMS/DRL references, P3/P4
locations). Based on the International 10–20 locations
(Myslobodsky, Coppola, Barziv, & Weinberger, 1990), channel names were: AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, AF4.
All physiological signals were then processed and statistically analyzed with custom software developed using MATLAB 7.10 (The Mathworks, Inc.; Natick, MA). Every channel was syn-chronously acquired at 2048 Hz and exported at a 256 Hz sampling rate (256 records per second, one every 3.90625 ms).
The lab was equipped with two portable PCs, one for recording the EEG and delivering the stimuli (with a 120 Hz display) and the other for psychophysiological signal recording.
To make interpretation relevant to actual users’ affective state and to avoid contaminations, light and temperature sensors have been used to monitor the conditions of the room and two three-axis accelerometer have been used to monitor subjects’ stability.
2.5. Signal processing
The collected data were analyzed using Matlab 7.10 (The Mathworks, Natick, MA). Cardiovascular and respiratory activity were monitored to evaluate both voluntary and autonomic effect of respiration on heart rate by analyzing R–R interval extracted from electrocardiogram and respiration (from chest strip sensor) and their interaction. Following the guidelines of Task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, typical Heart Rate Variability (HRV) spectral indexes were used to evaluate the
auto-nomic nervous system response (Camm et al., 1996; Magagnin
et al., 2010; Mauri et al., 2011). Spectral analysis was performed using Fourier spectral methods with custom software. The rhythms have been classified as very low frequency (VLF, <0.04 Hz), low fre-quency (LF, from 0.04 to 0.15 Hz), and high frefre-quency (HF, from 0.15 to 0.5 Hz) oscillations. This procedure enabled us to calculate the LF/HF ratio, also known as the sympathovagal balance index. Cardiovascular and respiratory activity interaction were assessed
using Respiratory Sinus Arrhythmia (RSA) index (Camm et al.,
1996; Magagnin et al., 2010). As temporal domain measure of heart rate variability, we calculated NN50 index, i.e., the number of inter-val differences of successive NN interinter-vals greater than 50 ms. This index describes the short-term NN variability. To simplify, NN intervals can be seen as a sort of beat-to-beat representation of heart rate. According to Camm et al. (1996), ‘‘In a continuous ECG record, each QRS complex is detected, and the so-called nor-mal-to-normal (NN) intervals (that is, all intervals between adja-cent QRS complexes resulting from sinus node depolarizations) or the instantaneous heart rate is determined.’’
The raw electromyography is a collection of positive and nega-tive electrical signals. Their frequency and amplitude give us infor-mation on the contraction or rest state of the muscle. Amplitude is measured in
l
V (micro-Volts). As the subject contracts the muscle, the number and amplitude of the lines increase, as the muscle relaxes, the number and amplitude of the lines increase decrease. We considered the Root Mean Square (RMS) to rectify the raw sig-nal and convert it to an amplitude envelope (Mauri et al., 2010). In this study, we were not interested in frequency related to muscle fatigue.Respiration signal has been measured to calculate the Respiration depth (RSP Depth), the point of maximum inspiration minus the point of maximum expiration was determined from the respiratory tracing. Smaller values indicate more shallow respiration and higher activation (Kunzmann, Kupperbusch, & Levenson, 2005; Mauri et al., 2010).
EEG signals were analyzed and filtered by independent compo-nent analysis (ICA) to remove ocular artifacts and blinks based on EOG signals through automatic algorithm and subsequent visual inspection. The corrected matrixes were then computed using spectral analyses to calculate 14 Slow Alpha EEG (7–10 Hz) bands, one per each channel recorded. Slow Alpha EEG bands have been
demonstrated to measure sustained attention (Klimesch, 1999;
Klimesch, Doppelmayr, Russegger, Pachinger, & Schwaiger, 1998). 2.6. Measures
Twenty psychophysiological measures were computed for the statistical analysis. In particular, 14 measured sustained attention, two measured physiological arousal, 2 measured emotional valence, and two measured anxiety. The measures considered will be following listed as follows: EMG Corrugator, EMG Zygomatic, HR, LF/HF, NN50, and Respiration Depth, extracted respectively from EMG-CS, EMG-Z, ECG, and RSP, following the mathematical procedures explained in signal processing. EEG measures consid-ered are named ‘‘Alpha – Ch. XX’’, where XX stands for the channel
name (based on the International 10–20 locations), and Alpha is intended for Slow Alpha EEG (7–10 Hz) bands. InTable 1we pro-vided the relationship between the psychological Variables and their psychophysiological measures (": is for increase; ;: is for decrease)
2.7. Experimental design
Data were analyzed by using repeated measure ANOVAs design with a 2 (PC vs. Mobile) 3 (Relax vs. Facebook vs. Stress). Since no effects emerged using different interfaces, we proceeded with two different analyses:
1. PC session. We conducted repeated measure ANOVAs design with the three conditions (Relax vs. Facebook vs. Stress) as within variable.
2. Mobile session. We conducted repeated measure ANOVAs design with the three conditions (Relax Mobile vs. Facebook Mobile vs. Stress Mobile) as within variable.
3. Results
Tables 2 and 5show the descriptive statistics respectively for PC and Mobile sessions: an increase of each dimension measured (physiological arousal, emotional valence, sustained attention and anxiety) corresponds to an increase or decrease in the effect of the variable used to measure it indicated through an up or down arrow respectively. The results of descriptive statistics reported herein showed that all the measure had the expected tendencies, for example, for physiological arousal, HR (up arrow, i.e., in the big-ger-is-higher form) resulted to increase from relax (low arousal), to Facebook (higher arousal) to stress (again higher arousal) while for Respiration dept (down arrow, i.e., in the smaller-is-higher form), it was opposite, since this measure is opposite respect to arousal (seeTable 4).
Repeated measures ANOVAs were performed with condition type (Relax, Facebook, Stress) as the within-subject variable for both PC and Mobile session, as reported inTables 3 and 6 respec-tively. The main significant effect of session type was found for most of the measures.
A pairwise comparison using the Bonferroni correction revealed significant statistical differences between Relax condition and Facebook condition and between Facebook condition and Stress condition.Fig. 2(PC setting) andFig. 3(Mobile setting) showed the graphical representation of some experimental measures reported in the tables (seeTable 7).
An EMG Corrugator similar to relax and a higher HR, indicate that during Facebook navigation, subjects have a positive emo-tional valence and an increased physiological arousal. Thus, sub-jects moving toward an engagement state and an optimal state during Facebook navigation is plausible. Furthermore, a lower LF/ HF index and an higher NN50 index confirm such hypothesis (Figs. 2 and 3).
Finally, polynomial a priori contrasts were performed by testing the hypothesized quadratic trends for the main effects in Physiological Arousal and Attention measures, with higher values for the Facebook and Stress conditions in comparison with the
Table 1
Description of variables and their measures (": is for increase; ;: is for decrease). Variable Measures
"Physiological arousal "Heart rate/"Skin conductance/"EEG beta waves "Emotional valence ;EMG corrugator/;EEG alpha asymmetry "Sustained attention ;EEG slow alpha waves
Relax condition in both PC and Mobile sessions (Tables 4 and 7). Polynomial a priori contrasts were also performed by testing the hypothesized quadratic trends for the main effects of Emotional Valence and Anxiety measures, with lower values for the Relax and Facebook conditions in comparison with the Stress condition in both PC and Mobile sessions. The statistical approach has been suggested by Cipresso and colleagues within a physiological tridi-mensional model of happiness (Cipresso, Serino, & Riva, 2014).
A monotonic trend, with measures increasing from the Relax to Facebook to Stress conditions, was also expected for both the PC and Mobile sessions.
Such analyses allowed us to get a deeper idea of affective states during Facebook navigation condition. Comparing the two condi-tions with relax and stress condicondi-tions, we can determine the effect of such navigation on the users in the Arousal-Valence-Attention tridimensional space.
Moreover, the analyses on polynomial contrasts allowed us to track the linear and quadratic trend within conditions, making it possible to track a conceptual map of the subjects’ affective states representation to understand the position of different conditions with respect to the theoretical optimal experience.
An interesting result, as also reported inFig. 4, indicated that Relax and Facebook conditions had similar EMG Corrugator (and also EMG Zygomatic) in PC session, i.e., they had similar positive emotional valence, although they had significantly different HR (and also Respiration Depth), with higher physiological arousal in Facebook condition. On the other hand, Facebook and Stress have similar HR (and Respiration Depth), i.e., similar physiological arou-sal (a little bit higher for stress, as easily understandable), but a sig-nificantly different EMG Corrugator (and also EMG Zygomatic), which suggests a positive emotional valence for Facebook condi-tion and negative for Stress.
Table 2
Mean (and St. Deviation) of the psychophysiological measures in the PC session.
Session type M (SD)
Measure Relax Facebook Stress
"Arousal HR " 69.819 (10.257) 74.689 (10.257) 79.876 (14.161) Respiration Depth ; 573.259 (209.646) 471.118 (208.874) 486.979 (181.106) "Valence EMG Zygomatic " 0.173 (0.096) 0.181 (0.094) 0.091 (0.047)
EMG Corrugator ; 0.088 (0.042) 0.089 (0.043) 0.186 (0.098) "Attention Alpha – Ch. AF3 ; 0.014 (0.012) 0.011 (0.011) 0.01 (0.008)
Alpha – Ch. F7 ; 0.006 (0.007) 0.006 (0.011) 0.005 (0.005) Alpha – Ch. F3 ; 0.019 (0.013) 0.011 (0.01) 0.012 (0.009) Alpha – Ch. FC5 ; 0.01 (0.009) 0.009 (0.01) 0.008 (0.007) Alpha – Ch. T7 ; 0.008 (0.007) 0.007 (0.005) 0.006 (0.006) Alpha – Ch. P7 ; 0.013 (0.01) 0.009 (0.009) 0.01 (0.009) Alpha – Ch. O1 ; 0.022 (0.016) 0.014 (0.012) 0.013 (0.009) Alpha – Ch. O2 ; 0.026 (0.017) 0.02 (0.016) 0.015 (0.012) Alpha – Ch. P8 ; 0.022 (0.02) 0.012 (0.01) 0.011 (0.011) Alpha – Ch. T8 ; 0.019 (0.017) 0.019 (0.02) 0.017 (0.015) Alpha – Ch. FC6 ; 0.012 (0.01) 0.009 (0.007) 0.009 (0.008) Alpha – Ch. F4 ; 0.017 (0.01) 0.012 (0.009) 0.015 (0.013) Alpha – Ch. F8 ; 0.007 (0.007) 0.005 (0.004) 0.007 (0.006) Alpha – Ch. AF4 ; 0.014 (0.01) 0.01 (0.008) 0.011 (0.009) "Anxiety LF/HF " 0.637 (0.265) 0.654 (0.287) 0.772 (0.232) NN50 ; 1407.318 (476.681) 2097.318 (1064.182) 976.364 (320.924) Table 3
Repeated Measure ANOVA and Pairwise Comparisons (Bonferroni correction) for the three conditions in the PC Session. Repeated measure ANOVA (Relax vs. Facebook
vs. Stress) PC session
Pairwise comparisons (Bonferroni correction) PC session
Measure F(48, 2) Sig. Partialg2
Facebook vs. Relax Facebook vs. Stress
Arousal HR 22.218 0.001 0.526 0.002 0.050
Respiration depth 8.752 0.001 0.285 0.006 n.s.
Valence EMG zygomatic 24.414 0.001 0.550 n.s. 0.001
EMG corrugator 33.279 0.001 0.625 n.s. 0.001
Attention Alpha – Ch. AF3 1.317 n.s. 0.052 n.s. n.s.
Alpha – Ch. F7 0.187 n.s. 0.008 ns. n.s. Alpha – Ch. F3 4.779 0.013 0.166 0.050 n.s. Alpha – Ch. FC5 0.774 n.s. 0.031 n.s. n.s. Alpha – Ch. T7 1.183 n.s. 0.047 n.s. n.s. Alpha – Ch. P7 2.803 0.071 0.105 n.s. n.s. Alpha – Ch. O1 8.214 0.001 0.255 0.037 n.s. Alpha – Ch. O2 6.474 0.003 0.212 n.s. n.s. Alpha – Ch. P8 6.883 0.01 0.223 0.012 n.s. Alpha – Ch. T8 0.294 n.s. 0.012 n.s. n.s. Alpha – Ch. FC6 2.069 n.s. 0.079 n.s. n.s. Alpha – Ch. F4 2.443 0.098 0.092 0.044 n.s. Alpha – Ch. F8 0.729 n.s. 0.029 n.s. n.s. Alpha – Ch. AF4 2.622 0.083 0.098 n.s. n.s. Anxiety LF/HF 1.761 0.185 0.081 n.s. n.s. NN50 16.609 0.001 0.454 0.025 0.001
In Mobile session, as can be noticed in Fig. 5, Relax and Facebook conditions showed similar psycho-physiological pat-terns, describing a positive emotional valence and a low physio-logical arousal (but lower in Relax). On the other hand, Facebook and stress conditions showed different psychophysiological pat-terns, as indicated by significant statistical differences in all cardio-vascular and electromyography signals.
This result needs to be examined deeper; in fact, the first analy-sis (valence-arousal model) could imply that Relax and Facebook are similar in the mobile setting, since they did not differ signifi-cantly in HR, although we expected signifisignifi-cantly higher HR for Facebook condition compared to PC session. Despite these similari-ties between Facebook and Relax in HR, a deeper analysis of
respiration depth and cardiovascular indices clearly differentiated between the two conditions. In particular, to enhance the results of the cardiovascular analysis, we examined LF/HF measure to try to study the dynamics of both sympathetic and parasympathetic activity. The results showed a lower LF/HF for Facebook mobile rather than Relax mobile (0.737 vs. 0.584), indicating a probable baroreflex effect in reducing HR. A clear variation was also found between Facebook and Relax condition in sympathovagal balance showing a state. Finally, a significant difference in respiration depth between the PC and the Mobile settings clearly indicates a high level of engagement, confirming our first research question.
To answer the second research question, we have to consider the results of all conditions. Based on the computed results, we
Table 4
Linear and Quadratic Contrasts in a Repeated Measure ANOVA for the three conditions in the PC Session. Repeated measure ANOVA (Relax vs. Facebook vs. Stress) linear contrasts – PC session
Repeated measure ANOVA (Relax vs. Facebook vs. Stress) quadratic contrasts – PC session Measure F(48, 2) Sig. Partialg2
F(40, 2) Sig. Partialg2
Arousal HR 34.359 0.001 0.632 0.022 n.s. 0.001
Respiration Depth 11.013 0.003 0.334 6.583 0.018 0.230
Valence EMG Zygomatic 22.329 0.001 0.528 30.732 0.001 0.606
EMG Corrugator 34.163 0.001 0.631 30.752 0.001 0.606
Attention Alpha – Ch. AF3 2.384 n.s. 0.090 0.102 n.s. 0.004
Alpha – Ch. F7 0.221 n.s. 0.009 0.176 n.s. 0.007 Alpha – Ch. F3 7.469 0.012 0.237 2.353 n.s. 0.089 Alpha – Ch. FC5 2.546 n.s. 0.096 0.001 n.s. 0.000 Alpha – Ch. T7 2.524 n.s. 0.095 0.098 n.s. 0.004 Alpha – Ch. P7 3.410 0.077 0.124 1.997 n.s. 0.077 Alpha – Ch. O1 18.835 0.000 0.440 1.640 n.s. 0.064 Alpha – Ch. O2 12.837 0.002 0.348 0.039 n.s. 0.002 Alpha – Ch. P8 6.365 0.019 0.210 9.887 0.004 0.292 Alpha – Ch. T8 0.627 n.s. 0.025 0.128 n.s. 0.005 Alpha – Ch. FC6 2.780 n.s. 0.104 1.289 n.s. 0.051 Alpha – Ch. F4 0.901 n.s. 0.036 4.425 0.046 0.156 Alpha – Ch. F8 0.021 n.s. 0.001 2.795 n.s. 0.104 Alpha – Ch. AF4 4.435 0.046 0.156 1.433 n.s. 0.056 Anxiety LF/HF 3.445 0.078 0.147 0.531 n.s. 0.026 NN50 20.954 0.001 0.512 16.035 0.001 0.445 Table 5
Mean (and St. Deviation) of the psychophysiological measures in the mobile session.
Session type M (SD)
Measure Relax mobile Facebook mobile Stress mobile
"Arousal HR " 70.22 (9.082) 72.066 (8.319) 80.801 (15.018) Respiration Depth ; 509.399 (189.137) 451.583 (180.546) 462.358 (188.677) "Valence EMG Zygomatic " 0.202 (0.104) 0.217 (0.11) 0.103 (0.107)
EMG Corrugator ; 0.101 (0.05) 0.113 (0.065) 0.227 (0.129) "Attention Alpha – Ch. AF3 ; 0.013 (0.013) 0.007 (0.006) 0.009 (0.008) Alpha – Ch. F7 ; 0.006 (0.008) 0.006 (0.008) 0.005 (0.007) Alpha – Ch. F3 ; 0.018 (0.016) 0.011 (0.014) 0.014 (0.012) Alpha – Ch. FC5 ; 0.012 (0.015) 0.006 (0.006) 0.009 (0.009) Alpha – Ch. T7 ; 0.011 (0.015) 0.005 (0.005) 0.008 (0.008) Alpha – Ch. P7 ; 0.015 (0.016) 0.007 (0.008) 0.011 (0.016) Alpha – Ch. O1 ; 0.023 (0.023) 0.011 (0.012) 0.017 (0.017) Alpha – Ch. O2 ; 0.028 (0.023) 0.012 (0.015) 0.021 (0.018) Alpha – Ch. P8 ; 0.022 (0.022) 0.009 (0.01) 0.017 (0.018) Alpha – Ch. T8 ; 0.022 (0.023) 0.013 (0.022) 0.018 (0.021) Alpha – Ch. FC6 ; 0.014 (0.02) 0.006 (0.006) 0.009 (0.01) Alpha – Ch. F4 ; 0.021 (0.019) 0.009 (0.009) 0.013 (0.013) Alpha – Ch. F8 ; 0.007 (0.01) 0.005 (0.005) 0.005 (0.007) Alpha – Ch. AF4 ; 0.018 (0.025) 0.015 (0.025) 0.012 (0.009) "Anxiety LF/HF " 0.737 (0.264) 0.586 (0.282) 1.033 (0.688) NN50 ; 1570.636 (762.253) 1569.091 (780.947) 518.318 (159.441)
illustrate inFig. 6the positions of these conditions in the tridimen-sional space defined by Arousal, Valence, and Attention.
Moreover, significant linear and quadratic contrasts indicated that emotional valence for Facebook condition is similar to Relax
but different from stress conditions, confirming positive emotion during Facebook condition, and that physiological arousal in Facebook condition is similar to Stress and different from Relax, confirming an higher activation during Facebook condition.
Table 6
Repeated measure ANOVA and pairwise comparisons (Bonferroni correction) for the three conditions in the mobile Session. Repeated measure ANOVA (Relax vs. Facebook
vs. Stress) mobile session
Pairwise comparisons (Bonferroni correction) mobile session
Measure F(48, 2) Sig. Partialg2
Facebook vs. Relax Facebook vs. Stress
Arousal HR 17.092 0.001 0.461 n.s. 0.002
Respiration Depth 5.960 0.005 0.221 0.012 n.s.
Valence EMG Zygomatic 17.490 0.001 0.467 n.s. 0.001
EMG Corrugator 28.805 0.001 0.590 n.s. 0.001
Attention Alpha – Ch. AF3 3.710 0.042 0.129 0.065 n.s.
Alpha – Ch. F7 0.552 n.s. 0.022 n.s. n.s. Alpha – Ch. F3 2.162 n.s. 0.080 n.s. n.s. Alpha – Ch. FC5 2.946 0.081 0.105 0.087 n.s. Alpha – Ch. T7 2.684 0.099 0.097 n.s. n.s. Alpha – Ch. P7 4.213 0.020 0.144 0.007 n.s. Alpha – Ch. O1 5.811 0.005 0.189 0.013 n.s. Alpha – Ch. O2 9.352 0.000 0.272 0.000 0.095 Alpha – Ch. P8 6.886 0.002 0.216 0.004 0.067 Alpha – Ch. T8 3.421 0.040 0.120 0.066 n.s. Alpha – Ch. FC6 4.683 0.032 0.158 0.036 n.s. Alpha – Ch. F4 7.059 0.002 0.220 0.010 n.s. Alpha – Ch. F8 1.066 n.s. 0.041 n.s. n.s. Alpha – Ch. AF4 0.684 n.s. 0.027 n.s. n.s. Anxiety LF/HF 6.562 0.010 0.247 n.s. 0.050 NN50 25.572 0.001 0.561 n.s. 0.001
Cardiovascular measures, namely NN50 in time domain and LF/ HF in frequency domain (spectral analysis), showed that Facebook condition did not produce stress.
Eventually, it seems that the Facebook navigation generated a higher physiological arousal. Moreover, cardiovascular measures,
EMG Corrugator, and EMG Zygomatic, clearly indicated subjects’ pleasantness during Facebook navigation. Thus, in general terms, Facebook is experienced as positive and activating, but not dis-turbing, gaining the subjects’ higher level of attention, as further demonstrated by the EEG Slow Alpha waves analysis.
Fig. 3. Mobile setting graphical representation of some measures.
Table 7
Linear and Quadratic Contrasts in a Repeated Measure ANOVA for the three conditions in the mobile Session. Repeated measure ANOVA (Relax vs. Facebook vs. Stress) linear contrasts – mobile session
Repeated measure ANOVA (Relax vs. Facebook vs. Stress) quadratic contrasts – mobile session Measure F(48, 2) Sig. Partialg2
F(40, 2) Sig. Partialg2
Arousal HR 19.338 0.001 0.492 9.329 0.006 0.318
Respiration Depth 4.885 0.038 0.189 8.641 0.008 0.292
Valence EMG Zygomatic 14.416 0.001 0.419 28.675 0.001 0.589
EMG Corrugator 30.646 0.001 0.605 22.746 0.001 0.532
Attention Alpha – Ch. AF3 2.585 n.s. 0.094 5.852 0.023 0.190
Alpha – Ch. F7 1.001 n.s. 0.039 0.000 n.s. 0.000 Alpha – Ch. F3 1.620 n.s. 0.061 2.475 n.s. 0.090 Alpha – Ch. FC5 1.011 n.s. 0.039 8.506 0.007 0.254 Alpha – Ch. T7 0.828 n.s. 0.032 6.833 0.015 0.215 Alpha – Ch. P7 1.263 n.s. 0.048 7.495 0.011 0.231 Alpha – Ch. O1 2.347 n.s. 0.086 9.318 0.005 0.272 Alpha – Ch. O2 4.026 0.056 0.139 16.067 0.000 0.391 Alpha – Ch. P8 1.655 n.s. 0.062 12.296 0.002 0.330 Alpha – Ch. T8 1.002 n.s. 0.039 8.010 0.009 0.243 Alpha – Ch. FC6 2.487 n.s. 0.090 11.978 0.002 0.324 Alpha – Ch. F4 6.131 0.020 0.197 8.089 0.009 0.244 Alpha – Ch. F8 0.865 n.s. 0.033 1.700 n.s. 0.064 Alpha – Ch. AF4 1.589 n.s. 0.060 0.003 n.s. 0.000 Anxiety LF/HF 3.717 0.068 0.157 13.782 0.001 0.408 NN50 43.230 0.001 0.684 11.616 0.003 0.367
3.1. Sensitivity and post hoc power analysis
A sample size calculation for the rmANOVAs yielded that approximately 28 subjects had to be used for this kind of experi-mental design in order to achieve a minimum power (1 b, where bis probability of Type II error) of 0.8 when considering a medium effect size of 0.25 and a significance level (
a
, wherea
is probability of Type I error) of 0.05. Once the results were computed, a power analysis was used to determine the likelihood that our study would yield significant effects.The goal of our post hoc power analysis was to compute achieved power, given the other three factors (the effect size, the sample size and
a
level), as indicated in the tables. The analysis showed a very good power, above 90% per each measure. Theseresults also demonstrated that our sample was appropriate (Faul, Erdfelder, Lang, & Buchner, 2007). However, this does not mean that our results can be easily generalized. This experiment was complex, and more studies are needed to make some further generalization.
Some results have been reported as significant at
a
levelbetween .05 and .10, since they achieved enough power to be rele-vant for the analysis. For this purpose, the values of partial-
g
2have been reported for the calculation of the effect sizes.4. Discussion
This study found statistically significant results regarding the psychophysiological correlates associated with navigating the
Fig. 4. Differences between Relax and Facebook and between Facebook and Stress for the PC session represented by valence-arousal model of physiological arousal and emotional valence through HR and EMG Corrugator, respectively.
Fig. 5. Differences between Relax and Facebook and between Facebook and Stress for the mobile session represented by valence-arousal model of physiological arousal and emotional valence through HR and EMG Corrugator, respectively.
Facebook. First, the comparison of Facebook condition with other conditions in both mobile settings and PC setting showed that experience, in general term, is different in many aspects of behav-ior as well as regarding the regulation of cardiovascular apparatus and of central nervous system. Concerning this, has been crucial to highlight the relevance of cardiovascular measures (Cacioppo et al., 2007; Rainville et al., 2006), as they offered us a more detailed information about all relax and stress states. Higher LF together with lower HF and lower NN50 clearly indicated a perceived stress in both the Mobile and PC cognitive tasks but not during Facebook navigation. Thus, it seems that the Facebook task did not stress subjects. To better understand this phenomenon, we considered these measures simultaneously rather than individually within a multidimensional model, such as the one included the physiologi-cal arousal, the emotional valence, and the sustained attention. Considering the engagement and the sense of presence during Facebook navigation simultaneously improved our analyses. Indeed, a higher physiological arousal could indicate both positive and negative affective states and emotions, depending on the attentional resources that subjects devote to the experience, which affect the sense of presence and the capability to have an optimal experience. For example, a higher physiological arousal could reflect anger, fear, age, or stress, and it can intensify with the nega-tive valence. On the other hand, the same phenomenon (higher arousal) could indicate joy, happiness, engagement, excitement, or ecstasy, and when associated with a high sense of presence, an optimal experience for the subject. In this sense, even with the bidimensional valence-arousal model, it is difficult to define subjects’ states. However, cardiovascular indices and electroen-cephalographic measures of attention added other interesting information to the model and made the results and our conjectures stronger than when considering just the two dimensions in the analysis.
With these considerations and limitations in mind, we sought to evaluate the experience of Facebook navigation and its capabil-ity to induce an optimal state in both PC and Mobile sessions.
According to these considerations, the experiment confirmed the hypothesis that psychophysiological measures describe a core Flow state during Facebook navigation through the PC, as also reported by Mauri and Colleagues (Mauri et al., 2011).
What clearly emerge is the strong capability of Facebook Mobile to generate the highest values across all dimensions, indicating
that it is the best inductor of presence and probably optimal experience, which is the most interesting challenge for future stud-ies and evaluations.
An element that led to the success of SNSs is their capability to enlarge and maintain social relationships, regardless of distance, and to strengthen close relationships. Although many researchers in the past have highlighted the lack of interpersonal media as
an effective alternative to FtF communication (Cathcart &
Gumpert, 1983; Jonassen & Kwon, 2001), recent evolution of tech-nology and SNSs demonstrated the opposite giving also evidences that communication strategies in FtF settings are often violated
and adopted more in mediated communications (Shuang-Shuang,
2010). Therefore, people tend to adopt communication strategies through the Facebook to strengthen their ties and efficiency in communicating. Social presence plays a key role in this process.
Social presence can be defined as the ‘‘sense of being with another’’ (Biocca, Harms, & Burgoon, 2003), in this case, it is related to the degree to which one feels another’s presence during Facebook navigation and in interpersonal-media interactions in general (Lee, 2004; Wood & Smith, 2001). The greater the band-width that an interpersonal medium can afford, the greater the social presence.
This is a key factor in our analysis. In fact, the hypothesis that a mobile could carry more cues is feasible, given present smart phones capabilities. Whether such cues are related to physical properties of such devices or to their ubiquitous capabilities remains to be verified.
5. Challenges and future studies
Rose is a clever and experienced researcher who travels around the world most of her time. She is visiting our lab and to wel-come her in Italy, we decided to take her around the city and dine together. At the end of the evening, we were very close and decided to take a photo together. Mike, one of our Colleague, has an amazing camera because of his great passion for professional photography, so he decided to set the self-timer and capture a great moment. Everyone knew that Mike would share a great photo on his Facebook profile. Rose asked every-one to wait because she wanted to take a picture with her iPhone 3GS, the older model with a common camera. Astonished, we explained to her that she would receive a great photo the next morning, as she would be tagged through Facebook. She explained to us that she is happy about this, but that something is missing in that great camera, the geo-lo-calization. Therefore, she opened a map in her mobile showing us all the places in the world where she took a picture and then explain to us the pleasure to have a location for each moment of her life, a pleasure that cannot be substituted by the quality of a photo. We learnt many things that night.
The story above helps us understand the role of ubiquitous technologies in enriching our life. Rose was happier to have a lower quality photos, but enriched by an information that – for her – was more important, that is, the context within which that photo was taken, her visiting a new country, and the memories of new friends (Fig. 7). In our opinion, its capability to improve people experiences and related affects makes a phone smart.
Future studies need to evaluate the experience of Facebook in mobile setting and consider ‘‘real mobile’’ setting, i.e., in vivo experiments conducted in the real world outside of the lab,
possi-bly using an ecological momentary assessment (Dockray et al.,
2010; Shiffman & Stone, 1998; Shiffman, Stone, & Hufford, 2008) ideally with the integration of wearable biosensors, as has already been done in mental health research (Gaggioli et al., 2011, 2012). The assessment of Facebook navigation in such a situation could
Fig. 6. A representation in the tridimensional space defined by physiological arousal, emotional valence, and attention.
greatly enhance the experience and further improve our knowl-edge of subjects’ experiences. In this study, we verified the device in a LAB setting. Therefore, future works should account for higher level of pervasiveness. De facto, only few studies have experi-mented with SNSs use in natural and daily environment. In fact, we found only one study that used a real mobile setting,
specifi-cally, the study of Humphreys (Humphreys, 2007) on
‘‘Dodgeball,’’ a mobile social network that has already been closed. Although the laboratory experiment maintains its status as the ‘‘gold standard’’ for controlled observation and precise testing of hypotheses, ubiquitous mobile platform offers the possibility to study user’s experience in real-time and in natural and daily environment. The use of a mobile phone increases also the ecologi-cal validity of the research.Bronfenbrenner (1979)defined ecologi-cal validity as ‘‘the extent to which the environment experienced by the participants in a scientific investigation has the properties it is supposed or assumed to have by the investigator.’’
To conclude, the consideration that arise is clear, i.e., we need to conduct the experiment outside the Lab, in real-life settings, to really understand the effect of ubiquitous technologies and their social use on Facebook.
Acknowledgements
This study has been made possible partially thanks to funds from European project, ‘‘INTERSTRESS – Interreality in the manage-ment and treatmanage-ment of stress-related disorders’’ (FP7-247685). References
Backs, R. W., da Silva, S. P., & Han, K. (2005). A comparison of younger and older adults’ Self-Assessment Manikin ratings of affective pictures. Experimental Aging Research, 31(4), 421–440.http://dx.doi.org/10.1080/03610730500206808. Barkhuus, L., & Polichar, V. E. (2011). Empowerment through seamfulness: Smart
phones in everyday life. Personal and Ubiquitous Computing, 15(6), 629–639. http://dx.doi.org/10.1007/s00779-010-0342-4.
Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory and measure of social presence: Review and suggested criteria. Presence-Teleoperators and Virtual Environments, 12(5), 456–480.
Blumenthal, T. D., Cuthbert, B. N., Filion, D. L., Hackley, S., Lipp, O. V., & Van Boxtel, A. (2005). Committee report: Guidelines for human startle eyeblink electromyographic studies. Psychophysiology, 42(1), 1–15. DOI 10.1111/j.1469-8986.2005.00271.x.
Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1).
Bradley, M. M., & Lang, P. J. (2000). Affective reactions to acoustic stimuli. Psychophysiology, 37(2), 204–215.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, Mass: Harvard University Press.
Cacioppo, J. T., Tassinary, L. G., & Berntson, G. G. (2007). Handbook of psychophysiology. Cambridge England; New York: Cambridge University Press. Camm, A. J., Malik, M., Bigger, J. T., Breithardt, G., Cerutti, S., Cohen, R. J., et al. (1996).
Heart rate variability – Standards of measurement, physiological interpretation, and clinical use. Circulation, 93(5), 1043–1065.
Cathcart, R., & Gumpert, G. (1983). Mediated interpersonal-communication – Toward a new typology. Quarterly Journal of Speech, 69(3), 267–277.
Cheung, C. M. K., Chiu, P. Y., & Lee, M. K. O. (2011). Online social networks: Why do students use Facebook? Computers in Human Behavior, 27(4), 1337–1343. DOI 10.1016/j.chb.2010.07.028.
Cipresso, P. (2011). Synchronizing physiological signals acquired from biofeedback equipment and eye-tracker systems. Applied Psychophysiology and Biofeedback, 36(1), 58.
Cipresso, P., Gaggioli, A., Serino, S., Cipresso, S., & Riva, G. (2012). How to create memorizable and strong passwords. Journal of Medical Internet Research, 14(1). ARTN e10, doi: 10.2196/jmir.1906.
Cipresso, P., Serino, S., & Riva, G. (2014). The pursuit of happiness measurement: A psychometric model based on psychophysiological correlates. The Scientific World Journal, 2014, 15.http://dx.doi.org/10.1155/2014/139128.
Cipresso, P., Serino, S., Villani, D., Repetto, C., Sellitti, L., Albani, G., et al. (2012). Is your phone so smart to affect your state? An exploratory study based on psychophysiological measures. Neurocomputing, 84, 23–30.http://dx.doi.org/ 10.1016/j.neucom.2011.12.027.
Codispoti, M., Surcinelli, P., & Baldaro, B. (2008). Watching emotional movies: Affective reactions and gender differences. International Journal of Psychophysiology, 69(2), 90–95. http://dx.doi.org/10.1016/j.ijpsycho.2008. 03.004.
Coyne, R. (2010). The tuning of place: Sociable spaces and pervasive digital media. Cambridge, Mass: MIT Press.
Dockray, S., Grant, N., Stone, A. A., Kahneman, D., Wardle, J., & Steptoe, A. (2010). A comparison of affect ratings obtained with ecological momentary assessment and the day reconstruction method. Social Indicators Research, 99(2), 269–283. http://dx.doi.org/10.1007/s11205-010-9578-7.
Draper, J. V., Kaber, D. B., & Usher, J. M. (1998). Telepresence. Human Factors, 40(3), 354–375.
El Emam, K., Moreau, K., & Jonker, E. (2011). How strong are passwords used to protect personal health information in clinical trials? Journal of Medical Internet Research, 13(1), 13–22. ARTN e18, doi: 10.2196/jmir.1335.
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G⁄Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191.
Favela, J., Tentori, M., & Gonzalez, V. M. (2010). Ecological validity and pervasiveness in the evaluation of ubiquitous computing technologies for health care. International Journal of Human–Computer Interaction, 26(5), 414–444. Pii 921368478, doi: 10.1080/10447311003719896..
Gaggioli, A., Cipresso, P., Serino, S., Pioggia, G., Tartarisco, G., Baldus, G., et al. (2012). An open source mobile platform for psychophysiological self tracking. Studies in Health Technology and Informatics, 173, 136–138.
Gaggioli, A., Pioggia, G., Tartarisco, G., Baldus, G., Corda, D., Cipresso, P., et al. (2011). A mobile data collection platform for mental health research. Personal and Ubiquitous Computing, 1–11. doi: citeulike-article-id:9866520.
Gruhn, D., & Scheibe, S. (2008). Age-related differences in valence and arousal ratings of pictures from the International Affective Picture System (IAPS): Do ratings become more extreme with age? Behavior Research Methods, 40(2), 512–521. doi: 10.3758/Brm.40.2.512.
Humphreys, L. (2007). Mobile social networks and social practice: A case study of dodgeball. Journal of Computer-Mediated Communication, 13(1).
Jonassen, D. H., & Kwon, H. I. (2001). Communication patterns in computer mediated versus face-to-face group problem solving. Etr&D-Educational Technology Research and Development, 49(1), 35–51.
Keil, A., Bradley, M. M., Hauk, O., Rockstroh, B., Elbert, T., & Lang, P. J. (2002). Large-scale neural correlates of affective picture processing. Psychophysiology, 39(5), 641–649.http://dx.doi.org/10.1017/S0048577202394162.
Klimesch, W. (1999). EEG alpha and theta oscillations reflect cognitive and memory performance: A review and analysis. Brain Research Reviews, 29(2–3), 169–195. Klimesch, W., Doppelmayr, M., Russegger, H., Pachinger, T., & Schwaiger, J. (1998). Induced alpha band power changes in the human EEG and attention. Neuroscience Letters, 244(2), 73–76.
Kunzmann, U., Kupperbusch, C. S., & Levenson, R. W. (2005). Behavioral inhibition and amplification during emotional arousal: A comparison of two age groups. Psychology and Aging, 20(1), 144–158. doi: 10.1037/0882-7974.20.1.144. Lang, P. J. (1995). The emotion probe – studies of motivation and attention.
American Psychologist, 50(5), 372–385. Fig. 7. An example of photo geo-localization in a mobile display.
Lee, K. M. (2004). Presence, explicated. Communication Theory, 14(1), 27–50. Magagnin, V., Mauri, M., Cipresso, P., Mainardi, L., Brown, E., Cerutti, S., et al. (2010).
Heart rate variability and respiratory sinus arrhythmia assessment of affective states by bivariate autoregressive spectral analysis. Computing in Cardiology, 37(5737930), 145–148.
Mauri, M., Cipresso, P., Balgera, A., Villamira, M., & Riva, G. (2011). Why is Facebook so successful? Psychophysiological measures describe a core flow state while using Facebook. Cyberpsychology Behavior and Social Networking, 14(12), 723–731.http://dx.doi.org/10.1089/cyber.2010.0377.
Mauri, M., Magagnin, V., Cipresso, P., Mainardi, L., Brown, E. N., Cerutti, S., et al. (2010). Psychophysiological signals associated with affective states. Annual International Conference of the Ieee Engineering in Medicine and Biology Society (Embc), 2010, 3563–3566.
Morris, J. D. (1995). Observations: SAM: The self-assessment manikin – An efficient cross-cultural measurement of emotional response. Journal of Advertising Research, 35(6), 63–68.
Myslobodsky, M. S., Coppola, R., Barziv, J., & Weinberger, D. R. (1990). Adequacy of the international 10–20 electrode system for computed neurophysiologic topography. Journal of Clinical Neurophysiology, 7(4), 507–518.
Orr, E. S., Ross, C., Simmering, M. G., Arseneault, J. M., & Orr, R. R. (2009). The influence of shyness on the use of Facebook in an undergraduate sample. Cyberpsychology & Behavior, 12(3), 337–340. http://dx.doi.org/10.1089/ cpb.2008.0214.
Rainville, P., Bechara, A., Naqvi, N., & Damasio, A. R. (2006). Basic emotions are associated with distinct patterns of cardiorespiratory activity. International Journal of Psychophysiology, 61(1), 5–18. http://dx.doi.org/10.1016/ j.ijpsycho.2005.10.024.
Ran, W., & Lo, V. H. (2006). Staying connected while on the move: Cell phone use and social connectedness. New Media & Society, 8(1), 53–72.http://dx.doi.org/ 10.1177/1461444806059870.
Riley, J. M., Kaber, D. B., & Draper, J. V. (2004). Situation awareness and attention allocation measures for quantifying telepresence experiences in teleoperation. Human Factors and Ergonomics in Manufacturing, 14(1), 51–67.http://dx.doi.org/ 10.1002/Hfm.10050.
Riva, G. (2010). I social network. Bologna: Il Mulino.
Riva, G., Davide, F., & Ijsselsteijn, W. A. (2003). Being there: Concepts, effects and measurements of user presence in synthetic environments. Amsterdam; Washington, D.C.: Tokyo: IOS Press; Ohmsha.
Riva, G., Wiederhold, B. K., & Cipresso, P. (2015a). The Psychology of Social Networking. Identity and Relationships in Online Communities (Vol. 2): De Gruyter.
Riva, G., Wiederhold, B. K., & Cipresso, P. (2015b). The Psychology of Social Networking. Personal Experience in Online Communities (Vol. 1): De Gruyter. Ross, C., Orr, E. S., Sisic, M., Arseneault, J. M., Simmering, M. G., & Orr, R. R. (2009).
Personality and motivations associated with Facebook use. Computers in Human Behavior, 25(2), 578–586.http://dx.doi.org/10.1016/j.chb.2008.12.024. Rubin, R. B., Rubin, A. M., Graham, E. E., Perse, E. M., & Seibold, D. R. (2009).
Self-assessment manikin. Communication Research Measures Ii: A Sourcebook, 336–341.
Russell, J. A. (1979). Affective space is bipolar. Journal of Personality and Social Psychology, 37(3), 345–356.
Salimpoor, V. N., Benovoy, M., Larcher, K., Dagher, A., & Zatorre, R. J. (2011). Anatomically distinct dopamine release during anticipation and experience of peak emotion to music. Nature Neuroscience, 14(2), U257–U355. doi: 10.1038/ Nn.2726.
Salimpoor, V. N., Benovoy, M., Longo, G., Cooperstock, J. R., & Zatorre, R. J. (2009). The rewarding aspects of music listening are related to degree of emotional arousal. PLoS ONE, 4(10). doi: Artn E7487 Doi 10.1371/Journal.Pone.0007487. Schupp, H. T., Stockburger, J., Codispoti, M., Junghofer, M., Weike, A. I., & Hamm, A.
O. (2007). Selective visual attention to emotion. Journal of Neuroscience, 27(5), 1082–1089. doi: 10.1523/Jneurosci.3223-06.2007.
Shiffman, S., & Stone, A. A. (1998). Ecological momentary assessment: A new tool for behavioral medicine research. Technology and Methods in Behavioral Medicine, 117–131.
Shiffman, S., Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1–32. doi: 10.1146/ annurev.clinpsy.3.022806.091415.
Shuang-Shuang, L. (2010). A tentative study of the impoliteness phenomenon in computer-mediated communication. Cross-Cultural Communication, 6(1), 92–107.
Tajadura-Jimenez, A., Pantelidou, G., Rebacz, P., Vastfjall, D., & Tsakiris, M. (2011). I-Space: The effects of emotional valence and source of music on interpersonal distance. PLoS ONE, 6(10). doi: 10.1371/journal.pone.0026083.
Wood, A. F. (2011). The tuning of place, sociable spaces and pervasive digital media. Journal of Communication, 61(1), E13–E15. http://dx.doi.org/10.1111/j.1460-2466.2010.01536.x.
Wood, A. F., & Smith, M. J. (2001). Online communication: Linking technology, identity, and culture. Mahwah (NJ): Lawrence Erlbaum Associates.
Zhou, T., & Lu, Y. B. (2011). The effect of interactivity on the flow experience of mobile commerce user. International Journal of Mobile Communications, 9(3), 225–242.
