First Monday

Investigating message forwarding behavior of mobile phone users: Exploring the link between message content, user sentiment, and user intention to forward messages on social media-based instant messaging platforms by Devendra Dilip Potnis, Bhakti Gala, and Kanchan Deosthali



Abstract
In an online survey, 108 mobile phone users in the age bracket of 18 to 21 in India reported their emotional responses to six humorous, warning, and philosophical messages in real time. Using open coding, researchers coded respondents’ sentiments into positive, negative, and neutral categories and traced their effect on (a) the respondents’ intention to forward the messages, which was captured in real time; and, (b) potential recipients of the forwarded messages. Findings inform the research on electronic word-of-mouth on social media-based instant messaging platforms and mobile phones. Implications in terms of identifying and containing the spread of misinformation on social media are discussed.

Contents

Introduction
Literature review
Methodology
Findings and discussion
Implications
Conclusion, limitations, & future research

 


 

Introduction

Social media platforms accessible over mobile devices enable billions of users across the world (Baulch, et al., 2020) to ubiquitously receive and forward billions of text, audio, and video messages every day (Matassi, et al., 2019). The ability to forward messages helps social media users and their contacts to stay in touch with each other, which in turn encourages message forwarding over social media (Cruz and Harindranath, 2020; Kaun and Schwarzenegger, 2014). However, the process of forwarding messages is not the same on all social media platforms, which is the motivation for undertaking this study.

Unlike Twitter or Facebook, social media-based instant messaging platforms like WhatsApp and Telegram require users to actively select one or more contacts stored on their devices (e.g., mobile phones) to intentionally direct forwarded messages to the selected contacts without their consent, which we label as message forwarding behavior (MFB). MFB depicts and reinforces information sharing on social media-based instant messaging platforms since whenever the user forwards the forwarded messages as is to others, they reinforce the message forwarding cycle on social networks (Nahon and Hemsley, 2013).

Gaps in and limitations of past research

  1. The following process is most common when forwarding messages over social media-based instant messaging platforms. (A) The user may passively receive messages from multiple contacts in their social network. Without actively seeking any information the user may come across pieces of information that they may or may not find relevant. (B) The user emotionally interacts with the message content and can experience a myriad of emotions, which is also known as user sentiment (Nahon and Hemsley, 2013). These emotions could be positive, negative, or neutral. (C) The user may take no action on the received messages or forwards them to a combination of personal and professional contacts, i.e., electronic word-of-mouth (eWOM). eWOM is defined as the act of spreading information over the Internet (Sanz-Blas, et al., 2017), and hence, represents information dissemination over the Internet. However, most of the existing research on information sharing on social media (e.g., Gruzd, et al., 2011; Loureiro, et al., 2014; Palka, et al., 2009; Stieglitz and Dang-Xuan, 2014; Thelwall, et al., 2010) does not treat this phenomenon as a continuous process with a series of dependent events, where the occurrence of one event (e.g., the receipt of a message) influences the occurrence of another event (e.g., the user’s decision to forward the message); instead it embraces a limited view of information sharing as a static, instantaneous, and ephemeral event (Toder-Alon and Brunel, 2018). Also, most of the above studies do not elicit the user’s emotional engagement with any specific message content in real time but instead rely on the user’s pre-existing opinion or experience. Hence, such research on information sharing cannot fully explain the MFB of users on social media-based instant messaging platforms.

  2. Past studies on information sharing focus on users of social media such as Twitter, weblogs, and picture-sharing online communities (e.g., Chen, et al., 2014), where they do not have to choose specific recipients from a list of contacts to share information. Information is mostly shared with and forwarded by strangers on Twitter (Hansen, et al., 2011), in contrast to the MFB of users over social media-based instant messaging platforms. Rarely any study asks users to select and report the specific types of social ties they would forward different types of messages to, which emulates the MFB on social media-based instant messaging platforms.

  3. Several studies examine the demographic, cognitive, psychological, organizational, and technological factors that influence a user’s intention to share information in the form of text, audio, video, and visual messages over social media (e.g., Aharony and Gazit, 2016; Anderson, 2016; Bowler, et al., 2018; Simon, et al., 2016; Potnis and Gala, 2017; Tandale, 2018; Thelwall and Vis, 2017) but rarely any of the past studies empirically investigate a theoretical link between message content, user sentiment and user intention to forward the message on social media-based instant messaging platforms. This link depicts the typical steps involved in the process of sharing information over social media-based instant messaging platforms.

  4. Past theories on information dissemination assume that information provider and information receiver are two distinct agencies (Duggan and Banwell, 2004), which does not necessarily hold for social media-based instant messaging platforms, where the receiver of the message also serves as the provider of the message when they forward the message to others. Thus, past theoretical understanding of information dissemination cannot be applied “as is” when studying message forwarding over social media-based instant messaging platforms.

  5. Blogs, publishers, journals, or librarians serve as information intermediaries in past research on information dissemination (Bar-Ilan, 2007; Saarti and Tuominen, 2017). In contrast, user sentiment, which results from the user’s interaction with the message content (Nahon and Hemsley, 2013), serves as an intermediate stage between the message content and the user’s intention to forward the message on social media-based instant messaging platforms. However, rarely any past research treats user sentiment as an intermediary for information dissemination or examines its effect on information dissemination over the Internet (i.e., eWOM).

Research goal

This study fills in all of the above gaps by examining the chain reaction between (a) the type of message content (e.g., humor, warning, philosophical) received by the user; (b) the types of emotions experienced by the user after reading the message in real time; (c) the user’s intention to forward the message; and, (d) the potential recipients of the forwarded message (see Figure 1). This study also contributes to the discourse on the meaning of the rise of social media-based instant messaging platforms like WhatsApp for digital humanities research (Baulch, et al., 2020) and civic engagement (Pang and Woo, 2020).

 

MFB of WhatsApp users: the linkage explored by this study
 
Figure 1: MFB of WhatsApp users: the linkage explored by this study. Note: Adapted from Nahon and Hemsley (2013).

 

 

++++++++++

Literature review

Message content and eWOM

Emotional words or emotional framing of messages can illicit user attention (Bayer, et al., 2012). An increased level of cognitive involvement can then lead to information sharing. After conducting an online experiment with 400 customers, De Keyzer, et al. (2017) reported that message valence exerts a significant impact on consumer eWOM responses, and the impact of message valence is stronger for hedonic services, as opposed to utilitarian services. Based on data sets of more than 165,000 tweets, Stieglitz and Dang-Xuan (2014) found that emotionally charged messages on Twitter, a communication platform serving as both a social network and a medium of information sharing (Hansen, et al., 2011), are retweeted quicker and more often than neutral ones. Findings from multiple studies on blogs, social networking sites like Twitter and online news portals suggest that positive and negative sentiments associated with messages heighten the attention and interest of users (Bayer, et al., 2012), which in turn influence their feedback (Huffaker, 2010; Stieglitz and Dang-Xuan, 2014), participation (Joyce and Kraut, 2006) and information sharing behavior (Berger and Milkman, 2012). Joyce and Kraut (2006) found that positive emotions in messages reinforce a sense of community and encourage continued participation; hence, most users share positive content more often than negative content that spreads conflict, damage, aggression, or failure.

Content that evokes physiological arousal or positive (awe) or negative (anger or anxiety) emotion is more likely to be forwarded (Dobele, et al., 2007), whereas content that evokes no arousal or emotion, like sadness, is less likely to be forwarded (Berger, 2011; Stieglitz and Dang-Xuan, 2014). For instance, rumors can spread due to the importance and ambiguity associated with them. Sentiment analysis of over 4.1 million tweets reveals that negative messages are reposted 1.6 times more rapidly and frequently than positive and neutral messages (Tsugawa and Ohsaki, 2015). Hansen, et al. (2011) found that only messages generating negative user sentiment in the news-segment are more likely to be shared than messages in the non-news segment that generate negative user sentiment. Hence, the first research question was as follows:

RQ1: Do users always forward emotionally charged messages?

User sentiments and message forwarding

Certain emotions, such as anger, anxiety, awe, or amusement, trigger physiological arousal, which in turn can drive information sharing (Berger, 2011). The greater the anxiety, the more the content of the rumor is important for the rumor recipient; as a result, rumors travel faster in high-anxiety groups than in low-anxiety ones (Anthony, 1973). Although positive content (e.g., videos classified as funny and cute by users) is more viral than negative content (e.g., videos full of anger and hatred), (Guadagno, et al., 2013; Gruzd, et al., 2011) the relationship between emotion and social transmission of messages is more complex than valence alone and is driven in part by arousal (Berger and Milkman, 2012). For instance, content that evokes either positive (awe) or negative (anger or anxiety) emotion characterized as high arousal is more viral. In support, their study on the sample of 6,956 articles published by the New York Times between 30 August and 30 November 2008, showed that articles generating emotion capable of arousing readers physiologically are more likely to be shared by readers over e-mail. Thelwall, et al. (2010) proposed an algorithm to predict the role of emotions in influencing a user’s intention to share information over social networking sites. Emotions of higher intensity, whether positive or negative, lead to higher transmission tendency.

RQ2: Do user sentiments (i.e., positive, negative, and neutral) always influence the user’s decision to forward messages?
     a. If not, what are the characteristics of messages whose transmissions are not influenced by user sentiments?

RQ3: Do user sentiments equally influence the user’s decision to forward messages with different types of content (e.g., humor, warning, philosophical) and valence (e.g., positive, negative, and neutral)?

It is important to note that most of the existing studies on this topic are based on the participants’ past behavior of forwarding messages, whereas this study asks study participants about their intention to forward messages.

User sentiments and message recipients

The act of forwarding messages is an example of interpersonal communication. According to Schutz’s (1991) fundamental interpersonal relations orientation (FIRO) theory, people engage in interpersonal communication for three reasons: inclusion (i.e., the need to be part of a group or the need for attention), affection (i.e., showing appreciation and concern for others) and control (i.e., the need to exert power in one’s social environment). Sentiments that align with or advance a combination of FIRO’s motivations of interpersonal communication are likely to prompt users to forward messages. After studying 582 undergraduate students at a university in a major metropolitan area in the U.S., Ho and Dempsey (2010) found that the following motivations induce youth to forward messages electronically: (1) the need to be part of a group; (2) the need to be individualistic; (3) the need to be altruistic; and, (4) the need for personal growth. Automated sentiment analysis of 46,000 tweets revealed that users who spread more negative than positive content are more prolific posters than positive users (Gruzd, et al., 2011). Also, users who argue the most on social media, to establish control over others, are more likely to receive content that generates negative sentiment (Ma, et al., 2014).

Users take pride in or enjoy sharing content that produces positive sentiment with strong ties, and vice versa (Baulch, et al., 2020). Guadagno, et al. (2013) discovered that anger-producing videos are more likely to be forwarded to weak ties or people who belong to rival groups, such as rival universities, whereas videos that spread happiness are more likely to be forwarded to close contacts. This finding is similar to the one revealed by Hansen, et al. (2011), where after studying thousands of tweets on climate change, they conclude that the message can go viral if it serves good news to the strong ties of the message originator or serves bad news to the public.

RQ4: Does the type of user sentiment (i.e., positive, negative, and neutral) influence the recipient type of forwarded messages?

 

++++++++++

Methodology

Data collection

One of the researchers announced this study at the Central University of Gujarat, India, for seeking participation from students above 18. The link to our semi-structured, online survey designed using Qualtrics was sent out to the students interested in participating in this study so that they can take the survey on any device (e.g., mobile phone, computer) of their choice. After filling in the informed consent form, study participants were displayed text messages and related questions sequentially (see the Appendix). One of the researchers had received these messages as forwarded messages over WhatsApp from his friends and family in India.

This study relied on text messages because the participants live in parts of India where low bandwidth and low-speed data rates typically inhibit mobile phone users’ ability to receive and forward audio and video messages. The Appendix presents the two humorous messages, two warnings, and two philosophical messages and related questions that survey respondents answered online. The participation of respondents in this study was voluntary. No incentives were offered in return for taking the survey. Participants were informed that their responses were confidential and that information from the questionnaire would be used for research purposes only.

The survey asked respondents to identify the strong and weak ties to whom they would forward messages. Based on past research (e.g., Guadagno, et al., 2013; Kozinets, et al., 2010), researchers displayed the following definitions of strong ties and weak ties to the study participants: strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions. Unlike past studies (e.g., Palka, et al., 2009) that automatically classify family and friends as strong ties, researchers provided a list of potential recipients to study participants and gave them the choice to classify them as strong ties or weak ties. For instance, each respondent had a choice to report childhood and family friends, a potential recipient of the forwarded message, as a choice with strong ties or weak ties.

The characteristics of the content and context of each message are described below.

Humorous messages

Two humorous messages (see the Appendix) in Hindi, the most widely spoken national language of India, were highly contextualized into contemporary Indian politics, judiciary systems, sports, and entertainment. Both messages are entertaining in nature, but also negative in terms of valence. They represent dark humor. Funnily, M2 refers to the injustice served by the court in response to a famous actor’s tragic murder of five people sleeping on a footpath in Mumbai. The content of M2 is also emotionally charged. Few studies test the effects of humorous messages with a negative valence on users.

Warning messages

Two warning messages (see the Appendix) provide information to readers; hence, researchers consider them informational with a negative valence. These long messages also have a high amount of information content. They are persuasion-oriented, hyped, relevant, useful, and carry social and emotionally charged information.

Philosophical messages

Philosophical messages like M5 and M6 (see see the Appendix) are forwarded frequently on social media-based instant messaging platforms like WhatsApp in India to provide a positive perspective of life.

Data analysis

Using open coding technique (Neuendorf, 2002), researchers coded and clustered user sentiments as qualitative data into three categories: positive, neutral, and negative sentiments (see “Message content → User sentiments” linkage section below). The classification of user responses into three sentiments is in line with LIWC (Linguistic Inquiry and Word Count, www.liwc.net) software, which uses a list of emotion bearing words to detect positive and negative emotion in the text (Thelwall, et al., 2010). Researchers traced the effect of each of these sentiments on user intention to forward messages (see “User sentiments → Intention to forward messages” linkage section below). Researchers calculated the percentages of potential message recipients and mapped them onto the explored linkage (see “User sentiments → Types of recipients” linkage section below). The inter-coder reliability between the two researchers was over 90 percent.

 

++++++++++

Findings and discussion

A total of 108 students in the age bracket of 18 to 21 participated in the study. Although there is a growing number of studies on the use of social media and/or mobile devices by young adults (Kaun and Schwarzenegger, 2014; Potnis, et al., 2017; Potnis, et al., 2018; Seo, et al., 2016) none of the existing research defines and studies MFB of young users (Pang and Woo, 2020). Two thirds of the participants in this study were men. Participants reported 15 languages as their native language and thus represent different parts of India. All of them owned a mobile phone.

The following subsections present findings that can be directly mapped onto the linkages depicted in Figure 1.

Message content → User sentiments” linkage

Table 1 indicates the percentages of positive, neutral, and negative sentiments created by three different types of messages. The two philosophical messages generated the least amount of negative sentiment among respondents. In contrast, the two warning messages made the majority of respondents experience a negative sentiment.

 

Table 1: Types of user sentiments
Note: Numbers indicate percentage values.
#PositiveNeutralNegativeTotal
M1602614100
M2244432100
M3431146100
M455936100
M59064100
M68497100

 

User sentiments → Intention to forward messages” linkage

Findings show that a positive sentiment for a humorous message does not necessarily encourage all respondents to forward the message. For instance, 60 percent of respondents experienced a positive sentiment for M1 but a little over half of them (i.e., 34 percent) would forward it to others (see Table 2). A positive sentiment generated by two humorous messages does not necessarily prompt most users to forward it. The negative and neutral sentiment for the two humorous messages leads to the same type of intention to act on the messages among most respondents.

 

Table 2: Intention to forward messages
Note: Yes: will forward the message; No: will not forward the message; Ot: other (e.g., maybe, not sure, or no action on messages); TT: total (numbers indicate percentage values).
 M1M2
SentimentYesNoOtTTYesNoOtTT
Positive3426060420024
Neutral12322612181444
Negative68014428032
 Total100Total100
 M3M4
SentimentYesNoOtTTYesNoOtTT
Positive37604346 9055
Neutral632116039
Negative2317646330336
 Total100Total100
 M5M6
SentimentYesNoOtTTYesNoOtTT
Positive701739070 9584
Neutral06067209
Negative04040707
 Total100Total100

 

A negative sentiment for warning messages does not always discourage all respondents from forwarding the message. The positive and the neutral sentiment for the warning messages leads to the same type of intention to act on the messages among most respondents.

In the case of philosophical messages, most respondents with a positive sentiment intend to forward them to others. Among most respondents, the positive sentiment for philosophical and warning messages leads to the same type of intention to act on the messages, but the same is not true for humorous messages. Similarly, for neutral and negative sentiments, the type of intention does not match for two messages of the same type.

An analysis of variance (ANOVA) test was conducted to compare the intention of study participants to forward M2, M3, and M4, the three emotionally charged messages. Results show that the F score (i.e., 5.98) for the “between-group variation” is greater than the F critical score of 3.02, with the p-value of 0.002 (see Table 3). The results suggest that there are significant differences in the participants’ intention to forward these messages. Hence, participants do not always forward emotionally charged messages.

 

Table 3: ANOVA results: Intention to forward emotionally charged messages?
Source of variationdf (degree of freedom)FP-valueF-CritSignificant difference?
Between groups25.980.002.023Yes
Within groups321    

 

Table 4 presents the results of the regression for the relationship between the user sentiment for each of the six messages and the intention of participants to forward each of these messages. Findings suggest that user sentiments influence their intention to forward all messages except M2. In the case of M2, the user sentiment does not influence their intention to forward this message (i.e., R-square = 0 percent, β = 0.004, p > 0.05). The values of R-square show that the degree of influence of user sentiments on their intention to forward messages varies for M1, M3, M4, M5, and M6. For instance, sentiments generated by M4 influence the intention to forward this message the most, since 42.19 percent of the variance in the intention to forward messages (p < 0.01) is explained by the sentiments experienced by users for this message.

 

Table 4: Regression results: Effects of user sentiments on the intention to forward messages
Note: p < 0.01
#R-square percentageCoefficient (β)Standard error
M15.410.160.068
M200.0040.07
M312.190.2310.06
M442.190.4280.048
M516.570.484.105
M68.95.292.09

 

Regression analysis of each user sentiment (i.e., positive, negative, and neutral) with the user intention to forward each of the messages shows that sentiments do not equally influence the user’s decision to forward messages with different types of content (e.g., humor, warning, philosophical) and valence (e.g., positive, negative and neutral).

User sentiments → Types of recipients” linkage

Young adults across the world use social media to manage their social ties (Baulch, et al., 2020; Seo, et al., 2016). Online communication on social media is useful for them to keep in touch with personal networks including friends and families (Kaun and Schwarzenegger, 2014). For instance, forwarding messages to others is one of the ways of maintaining social ties. Most respondents in this current study (a) with a positive sentiment for humorous messages would only forward them to childhood and college friends with whom they have strong social ties; and, (b) with a negative sentiment for humorous messages, would forward them to parents with whom they share strong or weak social ties. Parents are most likely to receive a forwarded message when their young children experience a negative sentiment from warning messages. Emotional bonding with parents and the subsequent concern for their safety seems like a driving force behind forwarding warning messages to parents. Figure 2 shows the role of sentiment in shaping user intention to forward messages to recipients with strong and weak ties.

 

Top three potential recipients of messages
Top three potential recipients of messages
 
Figure 2: Top three potential recipients of messages.

 

Table 5 presents ANOVA results of the types of recipients for each of the six messages when compared across three groups; namely positive, neutral, and negative sentiments. For M1, M2, M4, M5, and M6, the differences among the means of the types of recipients for the three groups of sentiments are statistically significant (p<0.005), i.e., the types of recipients of these messages varies significantly based on user sentiment. Thus, user sentiment influences the recipient type of five messages, irrespective of their content or valence.

 

Table 5: ANOVA results: Effects of user sentiments on types of recipients
#Source of variationdf (degree of freedom)FP-valueF-CritSignificant difference?
M1Between groups25.430.0043.02Yes
Within groups321    
M2Between groups234.50.003.02Yes
Within groups321    
M3Between groups21.630.193.02No
Within groups321    
M4Between groups211.550.003.02Yes
Within groups321    
M5Between groups2200.210.003.02Yes
Within groups321    
M6Between groups2103.840.003.02Yes
Within groups321    

 

Table 6 represents the percentage of strong and weak ties categorized according to the sentiments experienced by study participants for each of the six messages.

 

Table 6: User sentiments: Strong ties vs. weak ties
Note: Numbers indicate percentage values
#Strong tiesWeak tiesStrong tiesWeak tiesStrong tiesWeak tiesTotal
 Positive sentimentNeutral sentimentNegative sentiment 
M121.713.724.78.7238.2100
M22211.319.713.73.330100
M3303.3312.329.34100
M430.3333.3033.30100
M578220000100
M642833.516.500100

 

 

++++++++++

Implications

Theoretical contributions

This current study refutes several past findings. Theoretical contributions in Table 7 inform past research on the relationship between message content and message forwarding.

 

Table 7: User sentiments: Theoretical contributions
#Key findings of this current studyPast research comparisonsTheoretical contributions of the current study
RQ1There are statistically significant differences in the participants’ intention to forward M2, M3, and M4, three emotionally charged messages (see Table 3 above).Emotionally charged messages are shared more often than neutral ones (Stieglitz and Dang-Xuan, 2014).Not all emotionally charged messages are going to be shared by users since there are significant differences in the means of the intention of users to forward emotionally charged messages like M2, M3, and M4.
M2 generates anger against the existing legal system in India. However, the user sentiment does not influence the intention of users to forward it at all (see Table 4 above).Messages that trigger emotions, such as sadness, fear, humor, or inspiration, are more likely to be forwarded (Dobele, et al., 2007).The type of message content (i.e., humorous vs. warning) plays a key role in determining if emotionally charged messages are shared or not.
Emotionally charged warning messages with a high amount of informative content (i.e., M3 and M4) are going to be transmitted regardless of the emotions they generate among message receivers.Content that evokes physiological arousal or positive (awe) or negative (anger or anxiety) emotions is more likely to be forwarded (Berger and Milkman, 2012). 
RQ2M3 with a high amount of information is going to be shared by most of the users regardless of their sentiment for the message (see Table 2 above). In contrast, the user sentiment for M4, another message with a high amount of information, influences the intention of users to forward it the most (see Table 4 above).The amount of information embedded in message content prompts users to share the message (de Vries, et al., 2012).

The amount of information embedded in message content does not necessarily prompt users to share the message.

The user sentiment can hold more influence than that of the amount of information on the intention of users to share messages with a high amount of information.

The content of M3 and M4 is persuasion-oriented, hyped, relevant, useful, and carries social information. However, only M3 is going to be shared by most of the users regardless of their sentiment for the message (see Table 2 above).Message content that is persuasion-oriented, hyped, relevant, useful, or carries social information that builds individual reputation and group relationship is more likely to be shared with others (Kozinets, et al., 2010).User sentiments hold the greatest influence on the intention of users to forward M4 when compared across all six messages. This fact suggests that the message content that is persuasion-oriented, hyped, relevant, useful, or carries social information is not always going to be shared by users.

M3 and M4, two messages with informative content, are going to be shared by 66 percent and 55 percent of users, respectively, when compared to that of M1 and M2, two messages with entertaining content (see Table 2 above). Only 46 percent and 20 percent of users plan to share M1 and M2, respectively.

Philosophical messages like M5 and M6 are likely going to be shared by most users (i.e., 70 percent of users), followed by M3 and M4, the messages with informative content.

Entertaining content is shared more often than informative content by online communities (Jeon, et al., 2016; Luarn, et al., 2015).

Vs.
Informative content goes viral more often than any other type of content (Jenkins, 2011).

Entertaining content is not always shared more often than informative content.

Informative content does not always go viral when compared to philosophical content that shares wisdom and creates peace and harmony.

M3 and M4 represent the messages in the news-segment category. M3 is going to be shared by 23 percent of users, which represents the maximum number of users experiencing a negative sentiment among all messages. The overall influence of all the sentiments on the intention of users to forward M3 and M4 is 12.19 percent and 42.19 percent, respectively (see Table 4 above).Messages in the news-segment, which generate negative user sentiment, are more likely to be shared with others than messages in the non-news segment, which generates negative user sentiment (Hansen, et al., 2011).Messages in the news segment category, which generate a negative user sentiment, are more likely to be shared by users if the overall influence of all the sentiments on the intention of users to forward the message is low.
RQ3

Positive messages (i.e., M5 and M6) are likely to be forwarded more if they generate a positive user sentiment (see Table 2 above).

The spread of M1, an entertaining negative message, is influenced by user sentiments, whereas the spread of M2, another negative humorous message, is not influenced by user sentiments at all (see Table 4 above). Informative messages with a negative valence (i.e., M3 and M4) are likely to be shared regardless of user sentiment.

Message valence exerts a significant impact on consumer eWOM responses (De Keyzer, et al., 2017). Positive messages are shared more widely than negative messages. Most humans share positive content more often than negative content that spreads conflict, damage, aggression, or failure (Joyce and Kraut, 2006).

Vs.
Negative messages are shared 1.6 times more rapidly and frequently than positive and neutral messages (Tsugawa and Ohsaki, 2015).

Positive messages are not necessarily shared by all users. User sentiments influence the spread of positive messages.

A combination of user sentiment and message type (i.e., entertaining vs. informative) determines the spread of negative messages.

RQ4All six messages generate negative sentiments. However, users intend to forward all messages, except philosophical messages, to their strong ties (see Table 6 above).Messages that generate negative sentiment are not forwarded to strong ties in social networks (Guadagno, et al., 2013). A message can go viral if it serves good news to the strong ties of the sender (Hansen, et al., 2011).Only philosophical messages that generate negative sentiment are not shared with strong ties.
M2 is an anger-producing message. Fifty-five percent of users intend to forward it to their weak ties (see Table 6 above), irrespective of the sentiment experienced by them after reading this message.Anger-producing content is more likely to be forwarded to weak ties, whereas content that spreads happiness is more likely to be forwarded to close contacts or strong ties (Guadagno, et al., 2013).Anger-producing content is more likely to be forwarded to weak ties, irrespective of user sentiment.

 

Past studies show that positive sentiment is a strong predictor of eWOM. However, in this current study, findings related to philosophical messages alone support this finding. An overwhelming majority of respondents (i.e., over 80 percent of them) in this study experienced positive sentiment after reading both philosophical messages, and 70 percent of total respondents intended to forward both philosophical messages to others. Despite experiencing positive sentiment, only four percent of total respondents in this current study intended to forward M2, a humorous message, to others (see Table 2 above). This anger-producing message is political and questions the trustworthiness of the Indian legal system, which suggests that, in the case of messages on sensitive topics, message type holds a stronger influence than user sentiment on the user’s intention to forward.

This study challenges another finding from past research: messages that generate a negative sentiment are not forwarded to strong ties in social networks (Guadagno, et al., 2013). Study findings show that only philosophical messages that generate negative sentiment are not shared with strong ties (see Table 6 above).

In the case of warning messages like M3, user sentiment does not influence the type of recipient that the message generates. Regardless of the sentiment generated by warning messages, they are likely to be forwarded by most respondents in this current study. The top three potential recipients of warning messages for positive, neutral, and negative sentiments are strong ties (see Figure 2 above). Most of the responses indicating negative sentiment for warning messages centered on fear and worry. Respondents who felt scared, threatened, or worried after reading warning messages would forward the messages to parents, childhood friends and college friends with whom they have strong ties, so that they remain safe or protected from cyber threats and natural calamities. Thus, in the case of warning messages, user motivation seems to be more important than user sentiment in forwarding.

Recipients (e.g., colleagues at work, neighbors) sharing strong social ties with respondents are more likely than those with weak ties to receive all types of forwarded messages. Childhood and college friends with strong ties emerge as the most frequently reported category of recipients, regardless of message type and corresponding sentiment (see Figure 2 above). People holding strong social ties with youth are more likely to receive forwarded messages regardless of message content and subsequent user sentiment. This fact shows that, in the case of young mobile phone users, the strength of ties is more influential than user sentiment and message content. The recipient of the forwarded message does not necessarily depend on the message content or the corresponding user sentiment.

Most respondents intended to forward all messages that generated a positive feeling to the childhood and college friends with whom they share strong ties. This fact suggests that they care what the recipients would think about them (i.e., one’s social image), or that respondents think the recipients with whom they share strong ties would have a similar taste or liking for the specific message. Caring for the safety of childhood and college friends or a desire to create benefits for them are also likely motivations for respondents to share messages.

Practical implications

This study enriches MFB research with applications for businesses, governments, and society. More than four billion people worldwide use social media over mobile phones. In the background of the increasing popularity of social media and smartphones, there are several commercial and non-commercial applications and benefits associated with eWOM (Baulch, et al., 2020). Regardless of the time and location of citizens, governments spread the word about their policies, or create awareness among citizens with time-sensitive information (Cruz and Harindranath, 2020; Sandoval-Almazan and Kavanaugh, 2018). For instance, in the case of emergencies, governments, as part of mobile governance, broadcast information about escalated support and services to citizens via mobile phones, which could potentially change or save their life (Potnis and Gala, 2017). Mobile phone users expect their social contacts to be available daily and this expectation can become prominent in the case of emergencies (Ling, 2016). Social media-based instant messaging platforms like WhatsApp and Telegram, which are accessible over mobile phones, can play a key role in helping citizens spread the messages that they think are valuable for their social contacts.

Findings have multiple implications for government agencies and influencers on social media, especially for identifying and dealing with the spread of misinformation. First, findings suggest that fake news can be spread if the influencer composes emotionally charged warning messages with a high amount of content. The amount of content (i.e., high), the type of message content (i.e., warning), and the degree to which the message is emotional can be used as some of the criteria for identifying fake news. If a message meets one or more of these criteria, it is likely to be spreading fake news. Government agencies can apply these criteria to screen and identify fake news on social media-based instant messaging platforms like WhatsApp and Telegram. Governments can benefit from the fact that citizens are more likely to forward warning messages if the content of these messages is informative and emotionally charged.

Second, findings suggest that (a) messages in the news segment category, which generate a negative user sentiment, are more likely to be shared by users; and, (b) anger-producing content is more likely to be forwarded to weak ties, irrespective of the user sentiment. These findings can partly explain the spread of misinformation. For instance, the message that creates negative emotions such as fear, hate, and anger is likely to be forwarded by users. Social media users and influencers should verify the messages, especially those that generate a negative sentiment, before engaging with and forwarding them. Since messages that generate a positive sentiment are also likely to be false, it is advisable, in general, to verify the accuracy of the message content before forwarding them.

Finally, humorous messages that generate a negative sentiment are also likely to be forwarded by users to their strong ties. This finding suggests that government agencies can compose satirical messages to educate and inform citizens regarding public health emergencies such as COVID-19 and opioid crisis, among others; and citizens would still share them at least with their strong ties. For instance, often it is challenging to criticize religious practices. Messages that mock unhygienic religious practices that possibly promote the spread of COVID-19 are likely to generate a negative sentiment among users but would still be forwarded. Government agencies and businesses selling hand-sanitizers and similar products can benefit from our findings.

 

++++++++++

Conclusion, limitations, & future research

The empirical investigation of MFB in this study confirms the theoretical linkage between eWOM and information sharing research and situates eWOM as part of the MFB of social media users. Findings and theoretical contributions illustrate the significance of empirically investigating MFB as a process, an under-practiced approach in the research on information sharing over social media. This study proves that the effect of message content on message forwarding cannot be analyzed without considering the role of user sentiment, especially since user sentiment can be different than the emotion associated with the message content. This fact confirms user sentiments as a new intermediary in the existing research on information dissemination. Our study does not focus on features of any social media-based instant messaging platforms but the type of message content, user sentiments, and the user’s decision to forward messages; hence, study findings are generalizable and applicable to all social media-based instant messaging platforms.

A small sample size of messages, study participants from a narrow age bracket, and the sole use of text messages for the survey are some of the limitations of this study, which limits the generalizability of the research findings. In the future, researchers plan to compare findings from this current study with that of a new study involving a larger sample size of participants and a variety of messages, including audio and video messages. This study could have categorized user sentiments by employing the dimensional model of emotion from psychology (Russell, 1979), where sentiment is split into two axes: arousal (low to high) and valence (positive to negative). Nonetheless, researchers added neutral as a new dimension to the valence defined by Russell (1979) when categorizing user sentiments and studying their effects on intention to forward messages. This paper also did not employ any sentiment strength detection technique (Thelwall, et al., 2010) for scoring sentiments on any scale. In the future, researchers plan to study the effect of the source on MFB. Both humorous messages with a negative valence generated different sentiment responses, which also warrants an investigation of the role of the topic covered by the message, and the words of the message, in how it spreads. End of article

 

About the authors

Devendra Potnis is an Associate Professor at the University of Tennessee at Knoxville. His research focuses on the adoption of information tools, resources, and services by students, marginalized communities, libraries, microfinance, and governments. He has published his research in Communications of the AIS, First Monday, Government Information Quarterly, Information Development, Information Processing & Management, IT and Libraries, IT for Development, Journal of Education for Library and Information Science, Journal of Library & Information Science, LIS Research, Telematics and Informatics, The Information Society, and other reputed journals. He has received funding from the Institute of Museum and Library Services, OCLC, and ALISE.
E-mail: dpotnis [at] utk [dot] edu

Bhakti Gala is an Assistant Professor at the Central University of Gujarat, India. She obtained her doctoral degree in library and information science from the M.S. University of Baroda. Her research interests include digital libraries and the application of social media by libraries. She has presented research at over 10 national and international conferences. She is the winner of the 2014 International Paper Contest held by SIG-III, Association of Information Science and Technology. She has also served as a project manager for an international research project funded by OCLC and ALISE.
E-mail: : bhaktikgala [at] gmail [dot] com

Kanchan Deosthali is an Associate Professor of Management at the College of Business, University of Mary Washington. She obtained her doctoral degree in organizational behavior from the School of Business, University at Albany, State University of New York. Her research interests include citizenship behaviors, employee training and development activities, IT adoption by students and US small businesses, and e-book adoption. She has published her applied research in venues such as Journal of Business and Psychology, First Monday, Journal of Education for Library and Information Science, Southern Management Association, World Conference on e-learning, Association of the Global Management Studies, and Association for Information Science & Technology.
E-mail: : kdeostha [at] umw [dot] edu

 

References

Noa Aharony and Tali Gazit, 2016. “The importance of the Whatsapp family group: An exploratory analysis,” Aslib Journal of Information Management, volume 68, number 2, pp. 174–192.
doi: https://doi.org/10.1108/AJIM-09-2015-0142, accessed 21 July 2020.

Katie Elson Anderson, 2016. “Getting acquainted with social networks and apps: WhatsApp-ening with mobile instant messaging?” Library Hi Tech News, volume 33, number 6, pp. 11–15.
doi: https://doi.org/10.1108/LHTN-07-2016-0032, accessed 21 July 2020.

Susan Anthony, 1973. “Anxiety and rumor,” Journal of Social Psychology, volume 89, number 1, pp. 91–98.
doi: https://doi.org/10.1080/00224545.1973.9922572, accessed 21 July 2020.

Judit Bar-Ilan, 2007. “The use of Weblogs (blogs) by librarians and libraries to disseminate information,” Information Research, volume 12, number 4, at http://InformationR.net/ir/12-4/paper323.html, accessed 8 January 2020.

Emma Baulch, Ariadna Matamoros-Fernández, and Amelia Johns, 2020. “Introduction: Ten years of WhatsApp: The role of chat apps in the formation and mobilization of online publics,” First Monday, volume 25, number 1, at https://firstmonday.org/article/view/10412/8319, accessed 21 July 2020.
doi: http://dx.doi.org/10.5210/fm.v25i1.10412, accessed 21 July 2020.

Mareike Bayer, Werner Sommer, and Annekathrin Schacht, 2012. “Font size matters — Emotion and attention in cortical responses to written words,” PLoS ONE, volume 7, number 5, e36042.
doi: https://doi.org/10.1371/journal.pone.0036042, accessed 21 July 2020.

Jonah Berger, 2011. “Arousal increases social transmission of information,” Psychological Science, volume 22, number 7, pp. 891–893.
doi: https://doi.org/10.1177/0956797611413294, accessed 21 July 2020.

Jonah Berger and Katherine L. Milkman, 2012. “What makes online content viral?” Journal of Marketing Research, volume 49, number 2, pp. 192–205.
doi: https://doi.org/10.1509/jmr.10.0353, accessed 21 July 2020.

Leanne Bowler, Heidi Julien, and Leslie Haddon, 2018. “Exploring youth information-seeking behavior and mobile technologies through a secondary analysis of qualitative data,” Journal of Librarianship and Information Science, volume 50, number 3, pp. 322–331.
doi: https://doi.org/10.1177/0961000618769967, accessed 21 July 2020.

Jianqung Chen, Hong Xu, and Andrew B. Whinston, 2014. “Moderated online communities and quality of user-generated content moderated online communities and quality of user-generated content,” Journal of Management Information Systems, volume 8, number 2, pp. 237–268.
doi: https://doi.org/10.2753/MIS0742-1222280209, accessed 21 July 2020.

Edgar Cruz and Ramaswami Harindranath, 2020. “WhatsApp as ‘technology of life’: Reframing research agendas,” First Monday, volume 25, number 1, at http://dx.doi.org/10.5210/fm.v25i1.10405, accessed 21 July 2020.
doi: http://dx.doi.org/10.5210/fm.v25i1.10412, accessed 21 July 2020.

Freya De Keyzer, Nathelie Dens, and Patrick De Pelsmacker, 2017. “Don’t be so emotional! How tone of voice and service type affect the relationship between message valence and consumer responses to WOM in social media,” Online Information Review, volume 41, number 7, pp. 905–920.
doi: https://doi.org/10.1108/OIR-08-2016-0219, accessed 21 July 2020.

Lisette de Vries, Sonja Gensler, and Peter S.H. Leeflang, 2012. “Popularity of brand posts on brand fan pages: An investigation of the effects of social media marketing,” Journal of Interactive Marketing, volume 26, number 2, pp. 83–91.
doi: https://doi.org/10.1016/j.intmar.2012.01.003, accessed 21 July 2020.

Angela Dobele, Adam Lindgreen, Michael Beverland, Joelle Vanhamme, and Robert van Wijk, 2007. “Why pass on viral messages? Because they connect emotionally,” Business Horizons, volume 50, number 4, pp. 291–304.
doi: https://doi.org/10.1016/j.bushor.2007.01.004, accessed 21 July 2020.

Fiona Duggan and Linda Banwell, 2004. “Constructing a model of effective information dissemination in a crisis,” Information Research, volume 9, number 3, at http://InformationR.net/ir/9-3/paper178.html, accessed 18 January 2020.

Anatoliy Gruzd, Barry Wellman, and Yuri Takhteyev, 2011. “Imagining Twitter as an imagined community,” American Behavioral Scientist, volume 55, number 10, pp. 1,294–1,318.
doi: https://doi.org/10.1177/0002764211409378, accessed 21 July 2020.

Rosanna E. Guadagno, Nicole L. Muscanell, and David E. Pollio, 2013. “The homeless use Facebook?! Similarities of social network use between college students and homeless young adults,” Computers in Human Behavior, volume 29, number 1, pp. 86–89.
doi: https://doi.org/10.1016/j.chb.2012.07.019, accessed 21 July 2020.

Derek L. Hansen, Ben Schneiderman, and Marc A. Smith, 2011. Analyzing social media networks with NodeXL: Insights from a connected world. Burlington, Mass.: Morgan Kaufmann.
doi: https://doi.org/10.1016/C2009-0-64028-9, accessed 21 July 2020.

Jason Y.C. Ho and Melanie Dempsey, 2010. “Viral marketing: Motivations to forward online content,” Journal of Business Research, volume 63, numbers 9–10, pp. 1,000–1,006.
doi: https://doi.org/10.1016/j.jbusres.2008.08.010, accessed 21 July 2020.

David Huffaker, 2010. “Dimensions of leadership and social influence in online communities,” Human Communication Research, volume 36, number 4, pp. 593–617.
doi: https://doi.org/10.1111/j.1468-2958.2010.01390.x, accessed 21 July 2020.

Blaise Jenkins, 2011. “Consumer sharing of viral video advertisements: A look into message and creative strategy typologies and emotional content,” Master’s thesis in public communication, American University, at www.american.edu/soc/communication/upload/blaise-jenkins.pdf, accessed 3 December 2018.

Hyesun Jeon, Hyung Jun Ahn, and Gun Jea Yu, 2016. “What makes people react to the posts on the brand pages of mobile social network games?” Online Information Review, volume 40, number 3, pp. 435–448.
doi: https://doi.org/10.1108/OIR-07-2015-0236, accessed 21 July 2020.

Elisabeth Joyce and Robert E. Kraut, 2006. “Predicting continued participation in newsgroups,” Journal of Computer-Mediated Communication, volume 11, number 3, pp. 723–747.
doi: https://doi.org/10.1111/j.1083-6101.2006.00033.x, accessed 21 July 2020.

Anne Kaun and Christian Schwarzenegger, 2014. “‘No media, less life?’ Online disconnection in mediatized worlds,” First Monday, volume 19, number 11, at https://firstmonday.org/article/view/5497/4158, accessed 21 July 2020.
doi: http://dx.doi.org/10.5210/fm.v19i11.5497, accessed 21 July 2020.

Robert V. Kozinets, Kristene De Valck, Andrea C. Wojnicki, and Sarah J.S. Wilner, 2010. “Networked narratives: Understanding word-of-mouth marketing in online communities,” Journal of Marketing, volume 74, number 2, pp. 71–89.
doi: https://doi.org/10.1509/jm.74.2.71, accessed 21 July 2020.

Rich Ling, 2016. “Soft coercion: Reciprocal expectations of availability in the use of mobile communication,” First Monday, volume 21, number 9, at https://firstmonday.org/article/view/6814/5622, accessed 21 July 2020.
doi: https://doi.org/10.5210/fm.v21i9.6814, accessed 21 July 2020.

Sandra Maria Correia Loureiro, Hans Rüdiger Kaufmann, and Samuel Rabino, 2014. “Intentions to use and recommend to others: An empirical study of online banking practices in Portugal and Austria,” Online Information Review, volume 38, number 2, pp. 186–208.
doi: https://doi.org/10.1108/OIR-01-2012-0100, accessed 21 July 2020.

Pin Luarn, Yu-Fan Lin, and Yu-Ping Chiu, 2015. “Influence of Facebook brand-page posts on online engagement,” Online Information Review, volume 39, number 4, pp. 505–519.
doi: https://doi.org/10.1108/OIR-01-2015-0029, accessed 21 July 2020.

Long Ma, Chei Sian Lee, and Dion Hoe-Lian Goh, 2014. “Understanding news sharing in social media: An explanation from the diffusion of innovations theory,” Online Information Review, volume 38, number 5, pp. 598–615.
doi: https://doi.org/10.1108/OIR-10-2013-0239, accessed 21 July 2020.

Mora Matassi, Pablo J. Boczkowski, and Eugenia Mitchelstein, 2019. “Domesticating WhatsApp: Family, friends, work, and study in everyday communication,” New Media & Society, volume 21, number 10, pp. 2,183–2,200.
doi: https://doi.org/10.1177/1461444819841890, accessed 21 July 2020.

Karine Nahon and Jeff Hemsley, 2013. Going viral. Cambridge: Polity.

Kimberly A. Neuendorf, 2002. The content analysis guidebook. Thousand Oaks, Calif.: Sage.

Wolfgang Palka, Key Pousttchi, and Dietmar G. Wiedemann, 2009. “Mobile word-of-mouth — a grounded theory of mobile viral marketing,” Journal of Information Technology, volume 24, number 2, pp. 172–185.
doi: https://doi.org/10.1057/jit.2008.37, accessed 21 July 2020.

Natalie Pang and Yue Ting Woo, 2020. “What about WhatsApp? A systematic review of WhatsApp and its role in civic and political engagement,” First Monday, volume 25, number 1, at https://firstmonday.org/article/view/10417/8322, accessed 21 July 2020.
doi: http://dx.doi.org/10.5210/fm.v25i1.10417, accessed 21 July 2020.

Devendra Potnis and Bhakti Gala, 2017. “Factors influencing electronic word-of-mouth among Indian youth: Implications for mobile governance,” ICEGOV ’17: Proceedings of the Special Collection on eGovernment Innovations in India, pp. 107–114.
doi: https://doi.org/10.1145/3055219.3055241, accessed 21 July 2020.

Devendra Potnis, Kanchan Deosthali, and Janine Pino, 2017. “Investigating barriers to ‘using information’ in electronic resources: A study with e-book users,” Proceedings of the Association for Information Science and Technology, pp. 318–326.
doi: https://doi.org/10.1002/pra2.2017.14505401035, accessed 21 July 2020.

Devendra Potnis, Kanchan Deosthali, Xiaohua Zhu, and Rebecca McCusker, 2018. “Factors influencing undergraduate use of e-books: A mixed methods study,” Library & Information Science Research, volume 40, number 2, pp. 106–117.
doi: https://doi.org/10.1016/j.lisr.2018.06.001, accessed 21 July 2020.

James A. Russell, 1979. “Affective space is bipolar,” Journal of Personality and Social Psychology, volume 37, number 3, pp. 345–356.
doi: https://doi.org/10.1037/0022-3514.37.3.345, accessed 21 July 2020.

Rodrigo Sandoval-Almazan and Andrea Kavanaugh, 2018. “Introduction to the special issue on social media and government,” First Monday, volume 23, number 4, at https://firstmonday.org/article/view/8317/6659, accessed 21 July 2020.
doi: http://dx.doi.org/10.5210/fm.v23i4.8317, accessed 21 July 2020.

Silvia Sanz-Blas, Enrique Bigné, and Daniela Buzova, 2017. “m-WOM in a brand’s Facebook fan page,” Online Information Review, volume 41, number 7, pp. 936–953.
doi: https://doi.org/10.1108/OIR-08-2016-0237, accessed 21 July 2020.

William Schutz, 1991. “FIRO theory of needs,” In: Em Griffin (editor). A first look at communication theory. New York: McGraw-Hill, pp. 93–101.

Hyunjin Seo, Ren-Whei Harn, Husain Ebrahim, and José Aldana, 2016. “International students’ social media use and social adjustment,” First Monday, volume 21, number 11, at https://firstmonday.org/article/view/6880/5646, accessed 21 July 2020.
doi: https://doi.org/10.5210/fm.v21i11.6880, accessed 21 July 2020.

Tomer Simon, Avishay Goldberg, Dmitry Leykin, and Bruria Adini, 2016. “Kidnapping WhatsApp — Rumors during the search and rescue operation of three kidnapped youth,” Computers in Human Behavior, volume 64, pp. 183–190.
doi: https://doi.org/10.1016/j.chb.2016.06.058, accessed 21 July 2020.

Stefan Stieglitz and Linh Dang-Xuan, 2014. “Emotions and information diffusion in social media — Sentiment of microblogs and sharing behavior,” Journal of Management Information Systems, volume 29, number 4, pp. 217–248.
doi: https://doi.org/10.2753/MIS0742-1222290408, accessed 21 July 2020.

Manisha S. Tandale, 2018. “Use of WhatsApp as tool for information dissemination in the colleges of western region of Mumbai: A study,” International Journal of Information Dissemination and Technology, volume 8, number 3, pp. 147–149, and at https://www.ijidt.com/index.php/ijidt/article/view/8.3.5, accessed 21 July 2020.

Mike Thelwall and Farida Vis, 2017. “Gender and image sharing on Facebook, Twitter, Instagram, Snapchat and WhatsApp in the UK: Hobbying alone or filtering for friends?” Aslib Journal of Information Management, volume 69, number 6, pp. 702–720.
doi: https://doi.org/10.1108/AJIM-04-2017-0098, accessed 21 July 2020.

Mike Thelwall, Kevin Buckley, Georgios Paltoglou, Di Cai, and Arvid Kappas, 2010. “Sentiment in short strength detection informal text,” Journal of the American Society for Information Science and Technology, volume 61, number 12, pp. 2,544–2,558.
doi: https://doi.org/10.1002/asi.21416, accessed 21 July 2020.

Sho Tsugawa and Hiroyuki Ohsaki, 2015. “Negative messages spread rapidly and widely on social media,” COSN ’15: Proceedings of the 2015 ACM on Conference on Online Social Networks, pp. 151–160.
doi: https://doi.org/10.1145/2817946.2817962, accessed 21 July 2020.

 

Appendix

  1. Please select your gender:
        a. Male
        b. Female
        c. Other response: ___________
  2. Native language/mother tongue: ___________
  3. How long do you own a mobile phone?
        a. Less than 6 months
        b. Less than 1 year but more than 6 months
        c. Less than 2 years but more than 1 year
        d. Less than 4 years but more than 2 years
        e. Less than 6 years but more than 4 years
        f. Less than 8 years but more than 6 years
        g. Less than 10 years but more than 8 years
        h. More than 10 years

Disclaimer: The examples selected as forwarded messages in this study do not indicate researchers’ endorsement of these messages. Researchers do not encourage anybody to forward these messages. The messages were selected based on different categories of information.

Now let’s assume that you receive the following forwarded messages.

First humorous message (M1)

 

First humorous message

 

  1. How does this message make you feel?
    __________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    __________________________________________________________________________________________________________________________
  4. >If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

Second humorous message (M2)

 

Second humorous message

 

  1. How does this message make you feel?
    ______________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    ______________________________________________________________________________________________________________________________
  4. If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

Warning related to rainfall and possible help (M3)

 

First warning message

 

  1. How does this message make you feel?
    ______________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    ______________________________________________________________________________________________________________________________
  4. If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

Warning related to cybercrime (M4)

 

Second warning message

 

  1. How does this message make you feel?
    ______________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    ______________________________________________________________________________________________________________________________
  4. If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

First philosophical message (M5)

 

First philosophical message

 

  1. How does this message make you feel?
    ______________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    ______________________________________________________________________________________________________________________________
  4. If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

Second philosophical message (M6)

 

Second philosophical message

 

  1. How does this message make you feel?
    ______________________________________________________________________________________________________________________________
  2. Would you forward this message?
        a. Yes
        b. No
        c. Other response: _________________
  3. Why would you forward, not forward, or do something else with this message?
    ______________________________________________________________________________________________________________________________
  4. If you plan to forward this message, whom would you possibly forward it?
        a. Mother () Father ()
        b. Siblings with strong ties () weak ties ()
        c. Relatives with strong ties () weak ties ()
        d. Childhood Friends with strong ties () weak ties ()
        e. College Friends with strong ties () weak ties ()
        f. Colleagues with strong ties () weak ties ()
        g. Neighbors with strong ties () weak ties ()
        h. Teachers with strong ties () weak ties ()
        i. Hobby/Activity Group strong ties () weak ties ()
        j. Others: ________

Strong ties represent the people with whom you are frequently in touch or people who influence your decisions or actions; weak ties represent the people with whom you are not so frequently in touch or people who do not influence your decisions or actions.

 


Editorial history

Received 27 April 2020; revised 23 June 2020; accepted 10 July 2020.


Creative Commons License
“Investigating ‘message forwarding behavior’ of mobile phone users: Exploring the link between message content, user sentiment, and user intention to forward messages on social media-based instant messaging platforms” by Devendra Dilip Potnis, Bhakti Gala, and Kanchan Deosthali is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Investigating “message forwarding behavior” of mobile phone users: Exploring the link between message content, user sentiment, and user intention to forward messages on social media-based instant messaging platforms
by Devendra Dilip Potnis, Bhakti Gala, and Kanchan Deosthali.
First Monday, Volume 25, Number 8 - 3 August 2020
https://firstmonday.org/ojs/index.php/fm/article/download/10651/9587
doi: http://dx.doi.org/10.5210/fm.v25i8.10651