Share, Comment, and Like on Facebook and Message Strategies of Non-Governmental organizations

The scope of this study is to examine the relationship between message strategies and stakeholders’ responses on Facebook. Data on 20 non-governmental organizations’ Facebook pages was collected by simple random sampling to examine their communication practices on social media. It was found that those organizations were more likely to adopt Public Information (PI) model to transmit information to their stakeholders. Organizational messages based on Two-way symmetry model (TWS) can generate the greatest number of Likes which is significantly larger than Press Agentry (PA), Public Information (PI), and Two-way asymmetry (TWA) models. The multi-way (MW) model was proposed in order to extend Grung & Hunt’s framework in the digital era.


INTRODUCTION
Facebook has become the leading social media tool over the world and it has 2.5 billion monthly active users as of December 2019 (Statista, 2019).Studies have examined the communication methods of social media such as Facebook for public relations practitioners to maintain good contacts with stakeholders (Bonsón Ponte et al., 2015;Kim and Yang, 2017;Waters et al., 2009;Ihm, 2015).Charitable organizations and non-governmental organizations (NGOs) usually concern about their public relations budget.They try to keep their costs reasonable.As a result, social media could be one of the cheapest ways to reach most target audiences (Coombs and Holladay, 2012).
However, there is a lack of research attempting to examine what types of message strategies are used by organizations to communicate with their stakeholders via social media.Thus, the current study aims to investigate the types of message strategies that NGOs in Hong Kong use on Facebook and how these message strategies affect the level of stakeholder's engagement on their Facebook.

LITERATURE REVIEW
Facebook features include three different engagement tools (Share, Comment and Like) which are offered for people to respond to organizational messages.Two models have been proposed to measure stakeholder's engagement level on social media.Bonsón and Ratkai (2013) suggested a set of calculation metrics to measure popularity, commitment and virality for stakeholder engagement.Popularity relates to the number of Like, commitment is measured by Comment and virality is gauged by Share.This model calculates the responses from stakeholders by adding up popularity, commitment and virality of messages from people.The model assumed that there are same level of importance among Like, Share and Comment, because 'engagement' index is the sum of popularity, commitment and virality.There were usually different numbers of Likes, Shares and Comments in the same post, which indicates Likes, Shares and Comments require different efforts to respond to a particular post.
Organization message strategies used on Facebook could be examined by applying the "four models of public relations" initiated by Grunig and Hunt (1984) as a framework.Cho et al. (2014) proposed that different levels of engagement among stakeholders and organization could be captured by functions in Facebook, Like, Share and Comment, so as to embody each engagement feature's distinctiveness (Cho et al., 2014).Like is normally used on Facebook posts as an expression of pleasure or amusement  without any verbal or oral expression, which is the lowest level of engagement; Share need an effort to pass a message to somebody, which is the moderate level of engagement whereas Comment need more effort to react to organizational messages directly, it refers to the highest level of engagement.Grunig and Hunt (1984)'s "Four Models of Public Relations (FMPR)" are adopted in the current study since it covers both one-way and two-way communications (Figure 1).Public relations serve a propaganda function in the first model: Press agentry (PA).Public relations practitioners spread the preferred information by an organization.Regarding the public information model (PI), the purpose of public relations is to report objective information to stakeholders.For the two-way asymmetric public relations model (TWA), practitioners use researches and supporting figures to persuade stakeholders to accept the organization's opinion.As a result, people support the organization.In the two-way symmetric model (TWS), public relations people serve as a middle man among organizations and stakeholders.Their purpose is to foster understanding among the organizations and public.(Grunig and Hunt, 1984).

Hypotheses development
In previous studies, Kim and Yang (2017) investigated the relationship between organization messages and behaviour.Cho et al. (2014) employed the FMPR models to probe the non-profit organizational messages on Facebook.They reflected that NGOs are more likely to use public information model in organizational messages, which means NGOs mainly use social media to convey unilateral messages instead of fully utilize the interactive nature and dialogic function of the social media service (Waters and Jamal, 2011).By now, almost everyone is involved in social media sites to a certain extent, therefore, it is necessary to investigate the current situation.The first research question is presented as follows:

Research question 1:
To what extent do NGOs incorporate the FMPR on Facebook?Moreover, in order to examine whether there is an impact on stakeholder engagement between different message strategies among FMPR, the second research question is posited: Research question 2: How do message strategies affect levels of stakeholder engagement on these NGOs' Facebook pages?

Relevant hypotheses are listed as follows:
H1: There is a significant difference in the lowest level of stakeholder engagement (Likes) among "four types of organizational message strategies": Press Agentry(PA), Public Information (PI), Two-way asymmetry model (TWA) and Two-way symmetry model (TWS).
H2: There is a significant difference in the moderate level of stakeholder engagement (Shares) among "four types of organizational message strategies".H3: There is a significant difference in the moderate level of stakeholder engagement (Comments) among "four types of organizational message strategies".

Sampling procedure
A list of "Directory of Non-governmental Organizations" in Hong Kong was used as a sample frame (Community Engagement Discovery College, 2020).A random sampling was adopted to select 20 NGOs, each with an official Facebook page form the sampling frame.In accordance with the National Center for Chartable Statistics (NCCS), these NGOs can be categorized into 3 main types (that is, environmental and animal, health, and humanity).During 1 October to 31October 2019, there are a total of 443 posts uploaded on Facebook by these 20 NGOs in 31 days.All organizational postings will be coded for 1 month.All messages were collected in December 2019 and pasted to an Excel file, whereas the data were analyzed by SPSS software version 25in January 2020.The sampling size was determined by point of saturation.The researcher tried to collect more samples beyond the said period but the result was basically the same.

Coding procedure
A content analysis was employed.To code the NGOs' message strategies, this study adopted the coding schema based on Grunig and Hunt (1984).It was also developed and modified to explain the practices of FMPR on Facebook.Cho et al. (2014) used similar coding strategies.Codes were only assigned to the primary strategy if an organizational message involves multiple strategies.Moreover, three separate levels of stakeholders' engagement with postings uploaded by the above NGOs-the number of Likes, Shares and Comments on each post were coded.Additionally, the number of Likes of each organization's Facebook page and the number of posts within the data collection period were also coded.Two independent coders carried out the coding process.Thus the results were compared.In case of any discrepancy, discussion between them was allowed before they came to a final conclusion.Otherwise, a third coder was employed to examine the relevant posts and made the final decision.All the figures were cross-checked by a third independent person to maintain reliability and validity.

FINDINGS AND DISCUSSION
This study's sample included 10 environmental and animal NGOs, 7 human service NGOs, and 3 health NGOs.The number of posts in the October 2019 ranged from 1 to 89.The average number of post updated daily was 0.71.A total number of 1,267,384 Likes was recorded on the sampled NGOs' Facebook posts within the time period of coding, and the range was between 6,142 and 220,392.Thus, average number of Likes on a single post was 2860.A total number of 156,259 Likes was recorded on the sampled NGOs' Facebook posts within the time period of coding, and the range was between 30 and 111,438.Thus, average number of Likes on a single post would be 352.73.
Research question one examines to what extent NGOs incorporate the FMPR on Facebook.To answer this question, frequency counts were checked.TWA was the least used model of FMPR (N = 13 posts or 2.93%).The vast majority of the messages under this model were mainly used for promoting organizational events, persuading stakeholders to learn how to help or to be involved with the organizations.PI was the most predominant model (N = 284posts or 64.1%), followed by PA (N = 120posts or 27.0%).A greater part of PI model messages were providing updates and announcements from NGOs.The rest of the messages were sharing reports or information from other organizations.PA model was used by these NGOs to express organizational emotions.The TWS model (N= 26 or 5.87%) messages were mainly used to cultivate dialogue or give recognition and say "thank you" to supporters and/ or donors.Also, the function of hash tags was used in some messages to notify individual stakeholders about the updates on the Facebook page.
Research question two asked how organizational message strategies affect level of stakeholder engagement on these NGOs' Facebook pages.A series of one-way analysis of variance (ANOVA) test was carried out to test the hypotheses.Differences between each level of stakeholder engagement among FMPR have been found in this study.The ANOVA test analysis indicates that there is a significant difference in the number of the Like (lowest level of engagement) on a message in the FMPR (F = 8.091, p = <0.001).Thus, H1 is supported.From the Post hoc tests with multiple comparison, Model 4, two-way symmetry model (TWS) have a greatest number of Likes which is significant larger than the rest of the models.When compared with either PA (mean = 84.04,SD = 272.334),PI (mean = 120.43,SD = 266.893)or TWA models (mean = 100.77,SD = 150.905),stakeholders of these NGOs were more willing to Like Facebook messages based upon the TWS model (mean = 410.04,SD = 715.586).
The result is alike in the moderate level of engagement Share) on a message among four models ((F = 0.378, p = 0.769).As the p value are larger than 0.05, the null hypothesis cannot be rejected.Anon-significant difference is shown in the number of Comments (the highest level of engagement) (F = 0.303, p = 0.824) on organizational messages among four models as well.Thus, H2 and H3 were not supported.
According to the content analysis of NGOs' Facebook posts, the study found that PI model was the most frequently used model applied by NGOs, followed by PA, TWS and TWA.It was shown that the majority of NGOs use Facebook to promote the organizational events and transmit information to stakeholders.This result is contrary to the finding of previous studies (Cho et al., 2014;Waters and Jamal, 2011;Waters and Williams, 2011).It seems that NGOs in Hong Kong were less willing to engage stakeholders by using Facebook than NGOs in western countries.
The average of posts updated in the coding period was 0.71.In other words, most of the pages are updated twice by an organization every three days.According to Waters et al. (2011), organizations need to provide frequently update on their social media in order to ensure ongoing dialogue with their stakeholders.Therefore, it seems that a lot of researched NGOs have not fully utilized Facebook to engage with their stakeholders.This result is contrary to various studies.Many scholars found that lack of resources was the endured challenge for NGOs to actively utilize social media for relationship building (Cho et al., 2014;Briones et al., 2011;Nordström, 2012;Pavlovic et al., 2014).It would diverge some of organizations fans to the other sites.
The study showed that there was a significant difference in the lowest level of engagement (Like) and no significant difference in the highest level of engagement (Comment) among all four models.These results are contrary to the findings of Cho et al. (2014).Nevertheless, no significant difference exists in the moderate level of engagement (Share) among the FMPR which confirms the results of Cho et al. ( 2014)'s study.The finding shows that stakeholders of these NGOs were more likely to make Likes on TWS messages, in which NGOs were giving recognition to supporters, fostering dialogue, using direct messages to stakeholders with name tags.Despite the PI model was more frequently used by NGOs, stakeholders were not engaging with the messages based on this model.

Conclusion
This study investigated how NGOs use Facebook by revisiting the state of social media adoption in stakeholder engagement.Similar to previous studies, NGOs operating in Hong Kong is more likely to engage stakeholders by using Public Information communication model, which is basically one-way.TWA and TWS communication models are still rarely used by these NGOs, and even social networking sites are ideal platform in engaging people with organizational messages.
To better engage stakeholders by utilizing the FMPR message strategies on NGOs Facebook, suggestions are proposed as follows: First, at the earlier stage of establishing Facebook, it is suggested to use PA and PI to attract the stakeholders to become followers of the NGOs' Facebook pages.Several studies (Bonsón Ponte et al., 2015;Cho et al., 2014;Lovejoy and Saxton, 2012;Warren et al., 2014) proposed that information sharing "could be seen as core activity to attract" (Lovejoy and Saxton, 2012: 21) social media users.Second, TWA message strategy is recommended to attract people to get involved with the organization, such as joining organizational events, becoming donors or volunteers and giving specific feedback.Finally, it is better for NGOs to use TWS message strategy to achieve greater stakeholder engagement (Cho et al., 2014).
Facebook has a service in which one could buy Likes from target audiences for sponsor posts (Kim and Yang, 2017).Our findings might not show the true behaviour of respondents.It might be helpful for us to study stakeholder engagement using emoji, in which various small images are used to express the emotional attitude of the people without using text.
T. M. Wut* and C. Y. L. Lau, The Hong Kong Polytechnic University, Hong Kong.
Regarding the nature of communication, for the first two models, Press agentry and Public Information, they are always one way from the organization to public.In the twoway asymmetric model, practitioners plan what they want to communicate to stakeholders to achieve change in attitude and behaviour if possible.Communication from stakeholders is regarded as feedback only.Basically, the asymmetric model is a one-sided model.The true two-way communication model is the two-way symmetric model.It consists of dialogues among organizations and stakeholders.Stakeholders are likely to change organizational attitude and behaviour, whereas at the same time, organization persuades people to change their attitude and behaviour