Number of Comments Number of Users Percent Cumulative Percent
14 1 .09 .09
Table 4.4 cross-tabulates the stories by comment types. The overall percentages demonstrate that a full 70% of comments were monologic in nature. Twenty-one percent were responsive dialogue comments. Four percent were mutual discourse, while 5% of comments were of the responsive monologue type. There was significant cross-story variation in the type of comments made.
Table 4.4: Story * Type Cross tabulation
Story Type
‧
Because the types of comments and conversations varied so much across the different stories, it is useful to take a brief look at each individual story:
Haiti Relief: The comments on this story were overwhelmingly monologic in nature.
Almost 80% of the comments made on this story were user‟s opinions about Canada‟s planned relief work in Haiti. Like many of the comment threads at cbc.ca, replies were often political in nature either supporting or criticizing the current Conservative government. The little discussion that did take place in this thread was predominantly made up of political disagreements between users.
Tamils:The proportion of comment types left in relation to the Tamil Tiger story was similar to those on the Haitian Relief story. Many of the postings either rejected the two criminals in question as “Canadians” or made comparisons between the Canadian and American judicial systems and their respective sentencing records.10 This story did have a higher rate of responsive-monologue type comments. Many of these were criticisms of the at times xenophobic and exclusive nature of earlier commenters.
UFC Health:This story showed similar comment type trends as the earlier two. Much of the monologic content was aimed at insulting Mr. Lesnar the Ultimate Fighting Championship athlete in question, or defending Canadian health care systems. The higher rate of responsive monologue in comments related to this story is largely accounted for by users criticizing the perceived simplicity of posts attacking Lesnar and blindly supporting Canadian Health care.
10 Not long prior to this story, one of the “Toronto 18” terrorism plotters was sentenced to time already served plus one day. Many posters indicated a preference for the longer jail sentences of the American judicial system.
Quebecor Count 27 7 1 3 38
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Target: The story about Target‟s potential expansion into Canada facilitated significantly more (27.2%) dialogic content than most of the other stories. This is largely accounted for by users debating the quality of Target vis-à-vis competing department stores.
Some users argued for Target‟s superiority while others claimed that the potential expansion offered more of the same.
Polar Bears:The comment type breakdown of the Polar Bear export ban story was quite similar to the Target story. The debate between whether or not Northern communities such as Baffin Bay should be allowed to sell polar bear hunting licenses to – largely American – tourists garnered significant debate. The level of mutual discourse in this story was also higher than average as multiple users engaged in this debate and exchange levels increased.
Haiti Adoptees:The story about adopted Haitian children arriving in Ottawa attracted the largest proportions of dialogue and responsive monologue. The higher level of dialogue is largely accounted for by the numerous disagreements between users about adoption policies and national versus international priorities. Some users argued that Canadians should not adopt Haitians and should instead focus on Canadian children in need. These comments attracted many dissenting remarks resulting in a higher rate of dialogue. The higher rate of responsive monologue has a similar origin. Some of the comments arguing against adopting Haitians verged on xenophobia. This led some users to comment with a negative appraisal of the overall tone of the conversation.
Mumbai:While the story about the Mumbai terrorism suspect did not garner a huge number of comments, those it did attract are notable for the relatively high rate of mutual discourse. This is largely due to two more prolonged conversations in which users argued both about immigration policy and whether or not obtaining Canadian citizenship was enough to justify referring to an individual as “Canadian” and whether or not immigrants deserve to receive Canadian social welfare benefits.
Quebecor:The story about the American takeover of Quebecor‟s printing division attracted a fairly typical breakdown of comment types. The slightly higher rate of mutual discourse was the result of two conversations, one of which revolved around Quebec‟s allegiance to Canada and another which politicized the story and blamed the Harper Conservative government and NAFTA for Canadian economic woes.
‧
Asbestos:Jean Charest‟s visit to India and the potential for anti-asbestos protests attracted a high rate of monologic comments. The majority of these were criticisms of the perceived hypocrisy of Canada‟s ban on asbestos for domestic use while maintaining an export-oriented asbestos industry or the perceived hypocrisy of Jean Charest‟s environmental criticisms of Alberta‟s oil sands industry while he himself supports a potentially damaging industry.
Repatriation:The comments on the story regarding the repatriation of a Canadian soldier‟s body differed significantly from the average comment type distribution. The elevated levels of responsive dialogue and mutual discourse can be attributed to disagreements about the propriety of Canada‟s mission in Afghanistan and whether or not this particular news story offered an appropriate venue for discussion beyond sharing condolences with those involved in the story.
Olympic Flag: The story about Clara Hughes‟ appointment as Canada‟s flagbearer for the Olympic games opening ceremony did not elicit much dialogue. Rather, users tended to leave their opinions as to whether or not she was a good choice for the position. The small amount of dialogue that did take place mostly consisted of arguments about whether or not she was the best choice.
Feedback: Feedback was measured by tracking the number of users who clicked on the “Agree” and “Disagree” button linked to each comment. As such it is a variable that is most suited to assessing reactions to individual comments. However, looking at aggregate feedback data can give us some insight into agreement/disagreement behavior more generally.
In order to compare aggregate feedback levels, means were calculated by adding the total agrees and disagrees within various comment categories (i.e. story, comment type, address level) and then then dividing the total by the number of comments. Table 4.5 presents the agreement and disagreement means for each story. We can see that users are almost twice as likely to agree with comments as they are to disagree. Interestingly, the Haitian adoptee and the Polar Bear stories attracted the most balanced levels of agreement versus disagreement. It could be that this behaviorreflects the lack of consensus on the issues. The disagreements between users regarding both regional interests and the appropriateness of adopting Haitians may bereflected in contending levels of agreement and disagreement with the comments made. Conversely, a story like the one about Asbestos-related protests in India, which drew
‧
comments notable for their similar perspectives, attracted feedback that was overwhelmingly in agreement. All that said, the sample of 11 stories is not sufficient to draw final conclusions about what aggregate feedback data says about the nature of discourse within individual comment threads.
If we break down the feedback means by comment type as in Table 4.6 we see that Monologic comments including Monologue and Responsive Monologue types drew both more feedback and proportionally higher levels of disagreement than did dialogic type comments. Mutual discourse comments in particular drew fewer feedback clicks than did the other types. Perhaps as conversations between users became deeper, other readers not engaged in the conversation perceived them as somewhat tangential or private and subsequently felt less inclined to offer feedback.
Table 4.5: Agreement/Disagreement Means
‧
Responsive Dialogue Mean 30.48 12.69 2.40
Std. Deviation 40.967 17.757
Mutual Discourse Mean 20.54 7.07 2.91
Std. Deviation 19.046 8.683
Responsive Monologue Mean 36.25 15.67 2.31
Std. Deviation 35.105 17.889
Total Mean 36.56 17.24 2.12
Std. Deviation 49.124 28.863
Finally in terms of feedback we can see that, when broken down by address type as in Table 4.7, the micro-story component addresses drew a starkly higher proportion of agreement than did any other type of address. That is to say that when users made comments addressed at particular individuals mentioned within the stories, other users were much more likely to agree than to disagree with their comments. This may be due to the disproportionately high number of micro-story-component addresses within the UFC Health story. These comments, addressed to Brent Lesnar were almost universally negative in nature, and drew high rates of agreement. For instance, user Mrs. Jackatar wrote
“Unfortunately though in your country, you have to have $ to get the kind of care you received.” Seventy-five users agreed with this comment while only 5 disagreed.
Table 4.7: Address * Feedback
Address Agree Disagree Ratio
Macro-scopic Mean 37.13 19.05 1.95
Micro-story-component Mean 70.74 13.43 5.27
Std. Deviation 94.690 21.758
Total Mean 36.56 17.24 2.12
Std. Deviation 49.124 28.863
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
To summarize the answer to RQ1a, we see that an overwhelming majority of user interactivity at cbc.ca is expressive in nature. Only 25% of comments were dialogic (responsive dialogue or mutual discourse) and the vast majority of users only left one comment each. Using the „agree‟ and „disagee‟ buttons as a way to express feedback is popular with users, with agreement about twice as common as disagreement.