Hard News: Media Mathematics
133 Responses
First ←Older Page 1 2 3 4 5 6 Newer→ Last
-
Russell Brown, in reply to
I really don’t think you can blame Coleman for his interpretation of the figures because even if he DID calculate them incorrectly, the industry people quoting these figs should have known that Nielsens ( to the best of my knowledge) only produce monthlies!
Hi Jim!
But yeah, I think we can blame the office of the Minister of Broadcasting for coming up with a completely bogus to number try and back up a decision.
The odd thing is that, as I said, the daily number seems to be a real one. There aren't that many places it could have come from -- Juliet Jensen told me she'd never heard of it and she's the programmer.
-
Sacha, in reply to
But, hell, why should he have to try harder - and do better - when he has zero competition?
professional pride, ethics, and regulated standards?
-
Great article, I'm an interested observer from Australia. I think Currie is probably guilty of gross ignorance of audience estimation methods, not a sinister plot, but -as Lew Stoddart pointed out- the 'evidence' is presumably mere window dressing to serve a pre-determined goal.
I have minor quibble on precision of language, which I'd not raise except that journalists are (rightly) criticised here for reporting numbers badly and I think it's important.
Statistics expert Rachel Cunliffe offers 600,000 as a more accurate estimate than the invalid 207,000 figure for weekly cumulative audience, and Russell then states that:
... the audience figure that has come to be accepted as valid even by supporters of TVNZ 7’s retention was probably incorrect by a factor of about 200%.
This is not correct. It would be accurate to say that the false figure was 'probably incorrect by a factor of about 3'. You could also say that the actual audience was about 300% of the figure cited.
</pedantry>
-
Sacha, in reply to
I feel like we’re frogs that have been sitting in a pot of slowly boiling cluelessness.
nice
-
Russell Brown, in reply to
This is not correct. It would be accurate to say that the false figure was ‘probably incorrect by a factor of about 3’. You could also say that the actual audience was about 300% of the figure cited.
Fair call. I actually agonised over how to express that this morning. I think I'll have another go.
-
Graeme Edgeler, in reply to
I’m sure <grin> Graeme Edgler will correct me if I misquote him, but he makes the point that, while he knows NZ First got 4.07% of the party vote in 2008, he stills checks the figure every time before he writes it.
I generally only check when I'm quoting the exact number of votes: 95,356. Easy to end with a 96, because I'm usually also quoting the ACT number of votes: 85,496. I could work it out, as I remember the difference etc, but I figure people deserve more than that :-)
I also usually spell my name with five E's :-) But I almost never correct that, This seemed one of the rare occasions when it was appropriate, however.
-
Craig Ranapia, in reply to
But yeah, I think we can blame the office of the Minister of Broadcasting for coming up with a completely bogus to number try and back up a decision.
Sure - especially when, as Geoff Lealand and others could explain in great detail, the uses and abuses of media audience figures are (to use some academic jargon) problematic and highly contested.
Even if you give Coleman personally the benefit of the doubt (and I'm more inclined to than Russell, I suspect) it's not encouraging if a Minister is being given epically misleading information by staff. Of course, no Minister is obliged to accept the advice they're given, but it really helps if it's at least reality-adjacent. :)
-
Phil Lyth, in reply to
must remember, five
:-) :-) :-) :-) :-)
-
Phil Lyth, in reply to
At least you agonised.
Journos seems to have an innate capacity to make a mess of percentages. Typical is 'numbers have fallen 300%' when what they really mean is 'numbers are only 25% of the previous level'. Averages and other stats measures too.
-
Ben McNicoll, in reply to
reality-adjacent
I'm stealing that.
-
Well said, Russell. The fundamental point here is that the decision to pull the plug on TVNZ7 (which in my view has vastly superior content to all other NZ TV channels on average), was based on flawed data. This alone should provoke a review of that decision. If there were other, unstated arguments or criteria behind that decision, then let's see them please.
-
merc, in reply to
+1
-
izogi, in reply to
Journos seems to have an innate capacity to make a mess of percentages. Typical is ‘numbers have fallen 300%’ when what they really mean is ‘numbers are only 25% of the previous level’. Averages and other stats measures too..
It’s not just statistics and percentages. It’s simple multiplication. What does a phrase like “three times less than” even mean, given it combines multiplication by a positive whole number with a statement indicating a smaller value?
The intended message is probably something like “one third of” and the only literal interpretation I can think of is “-2 times the amount of”, but honestly, who knows?
-
decision... based on flawed data.
That's an assumption.
-
Craig Ranapia, in reply to
Journos seems to have an innate capacity to make a mess of percentages.
And we have a winner! Innumeracy is as big a factor here (if not more so) than some terrible political conspiracy; and there's nothing innate about statistical literacy. Like any other skill it has to be learned and maintained.
Of course, you can - and should - expect a lot better from senior journalists and ministerial staffers. (As a gratuitous aside, I'd also expect Coleman to be more than usually sensitive to not FUBARing numbers. In his previous life as a doctor, a miscalculated dosage could do pretty serious - or even fatal - harm.)
-
Martin Lindberg, in reply to
there's nothing innate about statistical literacy
Indeed not. I see people get percentage changes wrong all the time. Unfortunately the correct way is not intuitive (although simple), so you relly need to understand the basic formula.
-
As the risk of being a contrarian, all measurements of readership, viewership, listening are estimations and extrapolations of what much larger populations read, watch and listen to--and thus subject to all kinds of reservations, cautions and limitations. For example, the Nielsen ratings are based on self-reported measurements of presence in a room where a TV set is on, which might be something very different from watching television
-
all measurements of readership, viewership, listening are estimations and extrapolations of what much larger populations read, watch and listen to–and thus subject to all kinds of reservations, cautions and limitations
And tell us nothing about the level or quality of engagement with that material. A person may watch one hour of television per week, but that documentary or soap opera may be compelling and valuable to that person and enrich them greatly. Alternately, motor racing may play on a Sunday afternoon as background filler between salad and sausages. By way of comparison, I only read a few fiction books in a year, but they all mean something to me.
Now, we can dig in to these experiences with research about values, but simple metrics are by their very nature limited.
-
Euan Mason, in reply to
George Darroch, about an hour ago
ME: decision… based on flawed data.
GD: That’s an assumption.
From the original Herald story:
"Dr Coleman said the Government had no new money, and a channel that pulled in about 207,000 viewers a week - compared to the 600,000 people who tune in nightly to One News - was not a high priority. "
Unless the Herald has mucked up yet again, the assumption is valid.
-
merc,
The numbers determine advertising revenue, this is of the utmost importance to Mr Dr Coleman. If he misrepresents the viewer numbers he is also misrepresenting the revenue.
-
Craig Ranapia, in reply to
. For example, the Nielsen ratings are based on self-reported measurements of presence in a room where a TV set is on, which might be something very different from watching television
Sure - and they also don't reflect how people's consumption has fundamentally altered. Take Media7, for example, I very rarely watch the show "live" as opposed to a time-shifted viewing of the PVR recording, or catching a repeat over the weekend. Other people I know watch it On Demand.
Accurate or not, the media can also be very size queen-y about ratings (or movie grosses) without fairly reflecting that bigger isn't always better. One obvious commercial example: HBO shows routinely attract ratings that would be an automatic death sentence the "mainstream" networks. A subscription-based business model changes the game to a considerable degree.
-
Euan,
I believe that the decision was not based on the figures, but was an ideological decision, rationalised by the dodgy figures. -
Sorry that was me being a clever sod. The assumption is that Cabinet’s primary motive is poor ratings rather than an ideological position that the government has no role in providing television stations, and that these are better run and managed as for-profit private entities. Insomuch as this is the case, it wouldn’t matter if TVNZ7 gathered 2 million viewers per week. The only thing this would change is the political difficulty in executing such a decision. The relative ratings of 7 expedite this, in a way that Radio New Zealand is not yet subject to.
That ideological position can stand on its own, of course. There are arguments for and against. However, that’s an entirely different position, as indeed this post and your comments note.
[ETA: Now, I'm assuming that this is a Cabinet-driven decision and thus has Key, English, and Joyce behind it, rather than Coleman primarily.]
-
Craig Ranapia, in reply to
The numbers determine advertising revenue, this is of the utmost importance to Mr Dr Coleman. If he misrepresents the viewer numbers he is also misrepresenting the revenue.
Point of fact, Merc: Unless I'm seriously confused, TVNZ7 has no advertising revenue to misrepresent because it don't carry any ads other than internal promos. In this discussion, it would be wise to be more than usually careful about avoiding the very sin Coleman's being accused of.
-
Euan Mason, in reply to
Stewart
"Euan,
I believe that the decision was not based on the figures, but was an ideological decision, rationalised by the dodgy figures."Yes, You and George may well be right about this, but the best way to challenge ideology cross-dressed as rationality is to refute the dodgy rationality. After all, Coleman introduced his "rational" argument, not us.
Post your response…
This topic is closed.