Talk:Opinion polling for the 2018 Quebec general election

Proposed Deletion (June 2018)

Merge: No reason why this page should exist: very few Provincial elections have separate opinion polling pages, so unless all previous provincial general elections have opinion polling pages created the page should be merged with the Quebec general election, 2018 JDuggan101 (talk) 19:23, 21 June 2018 (UTC)[reply]
- Seconding JDuggan101's suggestion; there's no need to have this be a separate article. It is a tremendously useful compilation of polling information, and should be placed where it will get the most traffic. — Preceding unsigned comment added by HamartiaProsciuttoPharos (talk • contribs) 08:58, 3 December 2018 (UTC)[reply]

Quebec's larger importance in the federation is a good reason to keep this page. — Preceding unsigned comment added by 99.246.161.85 (talk) 02:26, 24 July 2018 (UTC)[reply]

Seat projections

I don't think we should be listing seat projections from so-called "poll aggregators" here. We already list the polls from which these projections are derived, and frankly anyone can come up with a black-box model to predict seat distributions from polling data; it doesn't make them a reliable or worthy source of information to add to Wikipedia. Cheers, Undermedia (talk) 12:14, 29 August 2018 (UTC)[reply]

P.S. I would further argue that TooCloseToCall probably falls into WP:SELFPUBLISH, as it's basically a personal website that's rarely if ever cited by the media. That being said, I'd even argue against including seat projections from poll aggregators hosted by media outlets such as the CBC's Poll Tracker, as they merely constitute "supplemental" analysis of original data that we already show on this page (i.e. the polls themselves). The pollsters themselves use well-established methods that conform to industry standards and were until very recently (the MRIA just recently called it quits) subject to industry oversight. The same cannot be said of poll aggregators and their seat projections. Undermedia (talk) 12:35, 29 August 2018 (UTC)[reply]

I believe it should be there, especially, CBC News, they are the poll trackers, they will be in charge of showing the votes, so what if, it’s aggregated, they are a broadcasting news, which is a perfect source. Ibelieve there should be seat projections, almost every single election article has this, for example, Opinion polling for the next Polish parliamentary election, Opinion polling for the next Slovak parliamentary election, etc. - Snowflakes Winter (talk) 1:57, 29 August 2018 (UTC)

I have removed these seat projections. One thing would be for pollsters to disclose their own seat projections, but projections based on poll aggregators have no sense, as we are not including poll aggregators. Impru20^talk 09:43, 31 August 2018 (UTC)[reply]

Polls missing exact date of polling can't be graphed

Just wanted to put that out there before someone asks why the pre-campaign period graph isn't being updated with the newly-added CROP polls going back to 2017 that only list the month of polling: the graph-generating code simply ignores them because they're missing data on the last day of polling. Cheers, Undermedia (talk) 11:42, 5 September 2018 (UTC)[reply]

It's ok to not graph them. But if you prefer to graph them, you could use the dates in the column "data collection" on this schedule. It is what I used to insert the lines at about the right places in the table. The dates when each poll was actually done may sometimes be slightly different from the dates planned in the schedule, for example the June 2018 poll was planned for June 13-18 but we see in the report that it was extended by one day to June 13-19. So, the dates in the planning schedule can be taken as a good approximation, knowing that they may have changed by a few days. In the table, I didn't put specific dates, but for the graph it should not make much difference if planned dates may differ from the actual poll dates by a few days. The schedule shows the planned dates from September 2017. Dates for the polls of April, May and June 2017 would still be missing, but usually the polls are near the third week of the month. -- Asclepias (talk) 14:47, 5 September 2018 (UTC)[reply]

COI concerns regarding Research Co.

@Undermedia: Hello. Regarding this edit, the link was added by 2001:569:7BB2:6C00:D4C7:9CE4:43F:71CE (talk · contribs · WHOIS). Based on behavior, this is probably operated by Research Co. IPs in this range (2001:569:7BB2:6C00:*) have been adding mentions from this firm or links to this website, and have not made any other edits at all. This appears to be spamming Wikipedia to promote this specific firm. For this article, since most of the entries in this list are from independently noteworthy polling firms, and this one is only supported by a primary source, I removed the listing as spam. If this specific study is significant, I will defer to your familiarity with the topic, but I would not assume, based on the current information here, that it is necessarily reliable, much less relevant. Thanks. Grayfell (talk) 23:10, 1 August 2019 (UTC)[reply]

Thanks for the explanation. I understand your concerns. From my perspective having contributed to opinion polling sections of many recent provincial and federal election pages, Research Co. appears to be a fairly recently founded polling firm that released its first political horserace polls during the 2018 Ontario election campaign before conducting these two polls during the 2018 Quebec election, subsequently polling during the recent Alberta election and most recently releasing its first federal horserace poll just last month. I'm aware the latter poll was reported in the Vancouver Sun in addition to Research Co.'s own website, and the firm's polls are being included in established poll aggregators such as the CBC's Poll Tracker and 338Canada.com, so they do appear to be legit. Cheers, Undermedia (talk) 00:09, 2 August 2019 (UTC)[reply]

Okay, that works for me. Thanks. Grayfell (talk) 06:25, 2 August 2019 (UTC)[reply]