Engagement behaviors across studied countries

Out of the total of 1914 reports, the active engagement behavior occurred in 536 i.e. consumers either clicked ‘like’/chose a reaction (463 times), shared (166 times), commented (108 times), or tagged (79 times) the company content they experienced. The most frequently occurring together behaviors were sharing and commenting content (Spearman correlation coefficient = .423, p < .01). In Table 21 I report the descriptive statistics on different engagement behaviors, and in Table 22 the differences between countries. The data featured 28% (536) cases where users engaged with the content and 72% of cases where users did not engage with the content (Chi²=370.41, p<.001).

Table 21. Engagement behaviors across studied countries

Engagement behaviors ∑ %¹

CLICKING LIKE 463 86.4 %²

SHARING 166 31.0 %

COMMENTING 108 20.1 %

TAGGING 79 14.7 %

TOTAL NO. OF ENGAGEMENT CASES 536 28.0 %³

NO. OF DIARY REPORTS 1914

1 They do not sum up to 100% as consumers could exhibit several active engagement behaviors at the same time e.g. both click ‘like’ and comment.

2 Read: 86.4% of diaries reporting active engagement involved clicking ‘like’ under the content.

3 The percentage of engagement cases in the total number of content reports made by diary participants – out of 1914 diary reports of experienced company social media content, consumers actively engaged with company content in 28% of cases (536).

The countries differed on the extent to which they engage with the content – for Finland, 8% of the diary reports involved active engagement behavior, in Poland 31%, and in US 37% (Chi² = 25.8, df = 2, p < .001).

The most common engagement behavior in all three countries was clicking ‘like’, which accounted for 86.3% (463) engagement reports, and there were no significant differences in the share of this behavior across countries. 31.0% (166) of the engagement behaviors involved sharing content, and the studied countries differed on this dimension (Chi² = 25.8, df = 2, p < .001), with almost 40% of the engagement behaviors in the U.S. involving sharing of the content, and around 19% in Finland and Poland. Countries also differed on the number of engagement behaviors involving commenting and tagging content. In the case of Finland, only 3% of the engagement behaviors involved tagging or commenting on the content, in the case of Poland 14% and 7% respectively, and in the case of the U.S. 25%

and 21% (commenting: Chi² = 13.9, df = 2, p < .01; tagging: Chi² = 22.0, df = 2, p

< .001). However, according to Cohen’s (1988) criteria, all those inter-country differences have a small effect, as the Phi coefficient is in the range between .10 and.30.

Table 22. Engagement behaviors across studied countries

Engagement behaviors

Country

Chi-square

statistics Phi Finland Poland USA

% OF ENGAGEMENT CASES¹ 7.9 % 30.9 % 36.6 % 125.83*** .256

% CLICKING LIKE² 84.0 % 89.4 % 84.5 % 2.53 .069

% SHARING² 18.8 % 19.6 % 39.7 % 25.80*** .216

% COMMENTING² 3.1 % 14.1 % 25.2 % 13.87** .161

% TAGGING² 3.1 % 7.1 % 21.0 % 21.96*** .203

Notes:

1 Read as the percentage of reports that reported active engagement in the total number of diary reports from a specific country

2 Read as the percentage of this type of engagement behavior in active engagement cases in a specific country; the percentages of different engagement behaviors do not sum up to 100% as one can click ‘like’, comment on, share or tag the same content at the same time.

*** p <.001; ** p < .01

The frequency of engagement varied across genders (Chi² = 6.25, df = 1, p < .05).

This association, however, has been minimal, with Phi = .057, well below Cohen’s (1988) criteria of .10 as having a small effect. Based on the odds ratio, women actively engaged with content 1.3 times more frequent than men. This can be attributed to the significant differences in commenting frequency (Chi² = 18.39, df = 1, p < .001), and sharing behaviors (Chi² = 6.23, df = 1, p < .05). In both cases the Phi below .10 indicates a minor effect. Based on the odds ratio, women commented on company content four times more frequently than men, and shared company content 1.6 times more frequently than men. However, there were no statistically significant differences between genders in tagging a friend (Chi² = 2.75, df = 1, p > .05), or clicking ‘like’/choosing a reaction (Chi² = 1.64, df

= 1, p > .05).

There were statistically significant differences between high and low in-group collectivism (Chi² = 4.16, df = 1, p < .05), however, the effect was much stronger when the groups of high and low assertiveness were compared (Chi² = 53.80, df = 1, p < .001). Based on the odds ratio, while cultures of high in-group collectivism have been engaging 1.4 times more frequently than cultures of low in-group collectivism, cultures of high assertiveness were engaging 2.1 times more frequently than cultures of low assertiveness. Cultures of high assertiveness were frequently engaging 1.2 times more than cultures of high in-group collectivism.

While sharing (in-group collectivism: Chi² = 9.22, df = 2, p < .01; assertiveness:

Chi² = 46.80, df = 2, p < .001), tagging a friend (in-group collectivism: Chi² = 7.55, df = 2, p < .01; assertiveness: Chi² = 61.88, df = 2, p < .001), clicking

‘like’/choosing reaction (in-group collectivism: Chi² = 6.54, df = 2, p < .05;

assertiveness: Chi² = 36.37, df = 2, p < .001) showed significant differences between the cultures scoring high and low on those dimensions. Interestingly, while for all the engagement behaviors, in the case of the cultural dimension of assertiveness, those effects were positive, in the case of sharing and tagging a friend, the effect was negative for the cultural dimension of in-group collectivism, i.e. the individuals from cultures scoring low for the in-group collectivism were more likely to manifest these engagement behaviors than in high in-group collectivist countries.

On the other hand, frequency of commenting behaviors did not differ based on the in-group collectivism score (Chi² = 1.70, df = 2, p >.05), but differed based on the assertiveness score (Chi² = 35.57, df = 2, p < .001).

In Table 22, the author reports descriptive statistics of the respondent motivations for content engagement in the three studied countries. The most frequently reported motive behind engagement behaviors were keeping in touch and self-expression (33%), followed by providing value (30%), self-presentation (21%), opinion leadership (13%), and expressing support (10%). As predicted, the percentages of the reported motivations driving engagement behaviors differed among studied countries.

The motive of keeping in touch is considerably more important in Poland and Finland, where 62% and 50% of the reported engagement behaviors, respectively, were motivated by this motive; with a relatively much lower percentage in the U.S. (13%). This difference is on a borderline medium/large effect (Cohen 1988) for the keeping in touch motive with a Phi coefficient of .497 (Chi² = 132.6, df = 2, p < .001).

A moderate effect of the country of origin exists on the self-expression motive as a driver of engagement behaviors (Chi² = 58.3, df = 2, p < .001). 46% of the engagement behaviors were motivated by self-expression in the U.S., and relatively less in Finland (22%), and Poland (14%).

The motive of providing value was a driver of 48% of the engagement behaviors in Poland, 34% in Finland, and 18% in the U.S. Thus, again, the extent to which these motivations are reported differs among the studied countries (Chi² = 50.9, df = 2, p < .001).

The self-presentation motive was most frequently reported in the U.S. sample (31%), compared to Finland (13%), and Poland (7%). These differences were also found to be significant (Chi² = 43.7, df = 2, p < .001).

Opinion leadership was most commonly reported as a motive-driving engagement behavior in the U.S. – it motivated 16% of the engagement behaviors, and one in ten behaviors in Poland and Finland. Expressing support was reported as a motive driving engagement behavior in Finland only in one case, and in 9% of the engagement behaviors in Poland, and 12.5% in the U.S.

Table 23. Motivations for engagement behaviors across studied countries

Motives ∑

Country Chi-square

statistics

Creamer’s Finland Poland USA V

KEEPING IN TOUCH 33.4 % 50.0 % 61.8 % 13.1 % 132.60*** .497***

SELF-EXPRESSION 32.8 % 21.9 % 14.1 % 46.2 % 58.33*** .330***

PROVIDING VALUE 30.0 % 34.4 % 47.7 % 18.0 % 50.87*** .308***

SELF-PRESENTATION 20.5 % 12.5 % 6.5 % 30.5 % 43.73*** .286***

OPINION LEADERSHIP¹ 13.4 % 9.4 % 10.1 % 16.1 % 4.23 .089

EXPRESSING SUPPORT¹ 10.4 % 3.1 % 8.5 % 12.5 % 3.93 .086 Notes:

*p < .05; **p < .01; ***p<.001

1 the minimum expected count in case of opinion leadership and expressing support is less than 5, thus violating the assumptions of square test. Therefore, the chi-square statistic is not reported here

Genders differed on the frequency of occurrence of the motives of keeping in touch (Chi² = 20.42, df = 1, p < .001), with Phi = -.195 standing for a small effect.

Based on the odds ratio, men are 2.4 times more likely to be motivated by the motive of keeping in touch than women. Significant difference was also found in the motive for self-expression (Chi² = 10.54, df = 1, p < .01), with Phi = .14 standing for a small effect. Based on the odds ratio, women are two times more likely to be motivated by the motive of self-expression than men. Genders differed also on the frequency of occurrence of the motives of expressing support (Chi² = 7.74, df = 1, p < .01), with Phi = .12 standing for a small effect. Based on the odds ratio, women are three times more likely to be motivated by the motive of expressing support than men.

There were no statistically significant differences between genders in the frequency of being motivated to engage by providing value to others (Chi² = .19, df = 1, p > .05), self-presentation (Chi² = 2.91, df = 1, p > .05), and opinion leadership (Chi² = .16, df = 1, p > .05).

5.3.2 Association between cultural dimensions and motives for