Latest RAJARs

[Russ]

What was the 'new methodology' introduced in 2007?

Russ

[french frank]

What was the 'new methodology' introduced in 2007?

Russ

Two things I remember were that they introduced new diaries which allowed people to specify the platform which they had been listening on; and there were changes to some of the TSA boundaries which affected local listening figures because different areas were being compared. I'm not sure how either would have affected overall listening, and in any case they did make it clear that, unlike with the 1999 change, figures before and after could be compared ('with care').

Edit: The details are here. I'm still not clear why the latest figures were a 'new record' under the new methodology. Ask Grant Goddard!

[Lateralthinking1]

I find RAJAR about as transparent as Opus Dei. This is peculiar given its area is communication. The sampling always appears ridiculously low, the information is spun as shown in the above article, I have no idea who the people running it are or who pays for them, and the general public are not permitted to see the detailed data. Is it in legislation a monopoly? I assume so in view of the fact that organisations like MORI don't do this stuff.

The independent radio consultant also makes an error. The gulf between pirate radio and the mainstream does not imply that the nation's youth will forever be disenfranchised. It is one of life's more bewildering features that old rock n rollers, mods, heavy rockers, punks, and e heads once found in a field off the M25, tend to end up listening to Ken Bruce. The radio habits change just as the clothes do. Goodness knows why.

It is obvious why the figures for Radio 3 have improved since last autumn. This coincided with a new, improved, forum.

[french frank]

Well, there is quite a lot of information about, Lat, and some of the answers are clear. As far as the sampling size goes, I would disagree with you: because samples are taken each week, there are 12 or 13 sets of figures which go into the quarterly figures and are presented as an average. If the weekly sample is c. 2,500 (which for the average opinion poll would already be quite high), the people questioned during the course of a quarter amounts to over 30,000 - a huge sample by most standards. And sampling progressively can iron out possible anomalies that might affect a survey taken on a single date.

The RAJAR website will tell you exactly 'who' RAJAR are: it is a limited company jointly owned by the BBC and RadioCentre, the umbrella group for all the commercial radio stations to provide a single set of verified figures which are accepted by all companies (except Kelvin MacKenzie's talkSport which always moans that its figures are too low and therefore wrong). Each company subscribes to RAJAR and in return they all get the same figures. One of the arguments against publishing all the figures is that if they did that some organisations wouldn't bother to subscribe at all, and RAJAR's revenue would be cut or it would increase the costs for the subscribers. But this only applies to the publication of all the data: it doesn't - and shouldn't - stop the BBC answering specific questions. We asked how many people listened to Choral Evensong when it was live on Wednesdays compared with how many listened when it was moved to Sunday afternoon; they said if they were to tell us that it might result in RAJAR being no longer financially viable. RAJAR, on the other hand, told us specifically that they had no interest in what the BBC did with its own figures and they were free to give us the information.

In fact, I think there is a more relevant point: RAJAR collects a huge amount of data, oceans of it. It requires special software to extract the individual pieces of information. It would be completely impossible to publish everything in a form the public could understand.

On monopoly: the figures would lose all value if there were five different companies all producing different figures: the whole point of RAJAR is that all companies (and, in the case of the commercials, all advertisers) should have the same figures.

I would tend to agree with you about Goddard's views on youth listening habits (though who can predict?): as long as there are radio stations somewhere out there pumping out what the kids like, they'll listen to radio. As they grow up, they'll listen to R2, R4 and R5. R1 can't be doing everything wrong as I believe they do get the biggest slice of the youth market.

[johnb]

The "RAJAR Key Facts" pdf document explains the methodology, etc in detail, should anyone be interested:

http://www.rajar.co.uk/content.php?page=about_key_facts

Then select "Click here to download the RAJAR Key Facts"

[Lateralthinking1]

Thank you french frank, and all other contributors, for your comments. I accept the point about sampling. It just doesn't come across in this way in the discussions. That RAJAR, the limited company, is jointly owned by both the BBC and RadioCentre should at least reduce the opportunity for bias towards either the BBC or commercial radio. However, that must mean that it is partially owned by licence payers. There is often a feeling that the BBC is less than accountable. Its payments to the well-known presenters spring to mind. I am therefore not sure that I am convinced by the RAJAR line that it is up to the BBC what it does with the figures. This seems to me a little like using the BBC's known reluctance to provide information as a smokescreen. In all bureaucracy now, there is that "it is isn't our area, it is theirs". The poor member of the public gets sent backwards and forwards like a tennis ball. Does this then really help the broadcasters, the advertisers, the listeners in every possible way?

I am not arguing that all the data should be published. However, it appears that there are restrictions on who can be subscribers. For example, those in broadcasting can be, universities probably can't be and members of the general public certainly can't be. Of the latter, only a very small number of people would apply if that were permitted. Enthusiasts who felt that they might be able to comprehend more, if not all, than is provided. It would appear from the links provided that Ipsos-MORI does have a significant role in the fieldwork.That is new to me. Would the figures really lack value if there were other organisations working in this area? This arrangement seems to work in terms of voting intentions with the potential for someone to provide an average. It would also enable a more independent process so that the broadcasters themselves were not directly involved.

My feeling is that the current system falls down on the "why" questions. Your question, frenchfrank, about why there should be an increase now in listeners to breakfast radio is a case in point. Political polling does place considerable emphasis on why people intend to vote for one party or another with a range of long-term policy indicators. I also note that RAJAR doesn't appear to have a sample of individuals who are interviewed throughout their lifetimes to see how and why their radio listening changes. And it isn't clear to me whether there is any significant research into the listening habits of many distinct groups, for example:

(a) ethnic groups
(b) those who live in this country for comparatively short periods
(c) people in the different regions - remember the concert survey for Scotland on the jazz boards?
(d) people in the cities, the suburbs, and rural areas
(e) those who prefer local radio and their expectations of it
(f) people who listen to a lot of radio and those who don't
(g) those who have a limited range of reception
(h) people with traditional radios, DAB sets, freeview etc
(i) those who prefer a tailored approach to radio and those who prefer a mix
(j) people who have been won over to the newer style stations and why
(k) people who like commercials and what they like about them
(l) people who tune in to pirate stations
(m) those who think radio is improving or getting worse

Finally, I agree from my limited experience that many move on to R5 as well as to R2, and some to R4. In fact, R5 could be the gateway into mainstream radio for many for a fair percentage of the pirate fans may also have an interest in sport. Lat.

[johnb]

From the "RAJAR Key Facts" document:

The following variables are available to RAJAR subscribers when analysing the data:
demographics
- sex
- age
- social grade
- ethnic origin
- region
- working status
- marital status
- household composition
- household tenure
- employment status

other media
- media access
- television viewing habits
- internet use
- mobile device use
- newspaper reading habits
- cinema attendance

Also the individual listening is detailed as:

At home
In car/van/lorry
At work/ Elsewhere

and

AM/FM
DAB
Digital TV
The Internet

[french frank]

I am therefore not sure that I am convinced by the RAJAR line that it is up to the BBC what it does with the figures.

Not quite accurate. The context was that the BBC suggested/implied/mumbled something about having a contract with RAJAR which 'limited' what they were allowed to disclose. When asked (by me ), RAJAR responded that this was not the case: there was no such contract and the BBC was free to do as it pleased. To quote them, 'Why should RAJAR care what they do with their own figures?'

On your 'why' questions: much of the information you mention is gathered by interviewers when they talk to survey participants. I was told by a now unfortunately departed member of BBC Audience Measurement that she would happily provide any information that they had 'off the shelf' which suggests that not all the information supplied by RAJAR is routinely processed. The latest story is that the BBC has a 'working practice' of disclosing details of listening 'when it is considered appropriate'. And I'm sure you can see the drawback in that ...

There is often a feeling that the BBC is less than accountable.

Oh! Wash your mouth out with soap and water, Lat! The BBC is worse than the Masons...

Oh, yes, subscribing to RAJAR: I was given to understand by RAJAR that this was just a question of money (and that FoR3 could have done so). But you have to subscribe for a year at least, it is expensive (at least a 4-figure sum), and, in our case, why would we want to shell out in advance for a welter of information which we don't have the ability to process and the vast majority of which we don't want anyway just in case there might be the odd detail we wanted to know?

[Lateralthinking1]

johnB - Yes, thank you for that clarification. I should have read the document fully and have now done so. Some of those items go further than my list. However, of those on my list, this list only applies to (a), (c) and (h). As an ordinary member of the public, I can only guess at the level of detail in the data for those categories but, with a few exceptions, it would appear that they are using fairly standard methodology. I had various things in mind when I drew up my list. As I mentioned, there is a distance between what people tend to do and why they do them. Certainly you could look at the categories in the "Key Facts" document and draw some conclusions - they were alone, with friends, at home, in the car, they prefer to be on the phone or at the cinema etc etc.

But what about - how have listening habits and expectations changed among the Asian groups between 1992 and 2011 and why?; how similar are the listening requirements of those from Afro-Caribbean groups over age 40 to (a) caucasians of the same age group and (b) those from Afro-Caribbean groups under age 40? - big changes there!; how have the listening expectations of those of ethic background as a whole changed in the last decade with a higher percentage of non-British Europeans?; can any distinct changes in radio listening in Scotland and Wales be detected since and because of devolution and is any trend there growing?

Or why do people tend to stick with one station in one listening session in contrast to TV viewing habits - is it to do with the design of sets, is this changing, would it be different with push button?; is there a greater sense of localism in some areas, is this reflected in radio listening, could the Localism Bill change these things?; would people prefer there to be more or less emphasis on celebrity presenters and what about payment?; to what extent do they value the traditions of the BBC?

And plenty more. I just think that there is more scope unless these things are being done and we are not hearing about them.

[Lateralthinking1]

....frenchfrank - Yes, that all sounds familiar. How British! I cannot begin to tell you how many times I have had the misfortune in the past six months......

"That's not Planning, it is Building Reguation" - "Oh no, it isn't us at Planning, it is Building Regulation, you will need to go back to them",

"Sorry, we don't deal with that at Ofcom, that is Otelo" - "Erm, wrong, I am afraid, not Otelo, it's definitely Ofcom but have you spoken to the phone company directly?"

"Water? That's not the Council, it is the water supplier" - "Is that what they said? No, it isn't us, it is for the residents to resolve, oh, hold on, the Government is telling us we might take it all over in October".

Beyond this pantomime of torture, I do accept that the cost for FoR3 of obtaining the detailed data is prohibitive, particularly as we only want to know bits and pieces. I sound like I half-understand these things. I don't really but I specialized many years ago in voting behaviour. I am always peculiarly interested in the whys and wherefores. It just seems to me that some additional depth, or at least more nuance, could lead both to improvements in broadcasting and commercial viability.

[Russ]

2011 Q2 results are now published. Compared to the excellent Q1, the Q2 results for Radio 3 are slightly disappointing: reach is down 3.8% at 2174k, and total hours are down 3.2% at 13362k. (And in case you might be wondering about the CFM 'competition', CFM had a rather miserable Q2, with its reach and total hours down about 6%, although bear in mind the CFM figures dwarf those of Radio 3.) Overall year-on-year trend for Radio 3 is still upward: 2010 averages were 2061k (reach) and 11842k (total hours per quarter).

Russ

[amateur51]

Cheers, Russ - nicely informative

[french frank]

Cheers, Russ - nicely informative

Yes, indeed, nice to have someone else to do the various calculations .

One or two comments: Radio 3 has been on an upward trend for a number of quarters now having hit rock bottom in about June 2007. The June quarter can sometimes be rather lower (and a lot of stations have slightly lower figures this quarter - R4 one exception). Because R3 hit a near record figure last quarter, I half expected the new live concerts which started in May (that's 8 of the 13 weeks), plus the connected publicity, to push it up again to a new record, but it stayed back.

The near 'sensational' figure which the BBC chose to highlight was a year-on-year 17% rise in reach. Impressive. But only because the June figure last year was unusually low (only 1.858m) so the bounce -back from the temporary low was good - but not as good as a 17% rise above an average figure!

On Breakfast, I make no assumptions. The programme maintained its excellent reach of last quarter, dropping back only slightly. It has made huge strides in the last two years while Classic FM's programme is faltering a bit ... the picture is not quite clear because CFM has altered its breakfast arrangements recently.

Edit: On CFM's drop in reach - the March quarter was very good (by recent standards), the first time they had passed 6m for ages. But they haven't maintained it and are back to a not disastrous but rather lack-lustre performance.

[johnb]

I've updated (hopefully correctly) the chart that I posted for the first quarter.

As one can see, there is usually (but not always) a fall back in the second quarter so this quarter's results are merely following that pattern.

[Russ]

As a reminder of the R3 2011 Q3 results (keeping the RAJAR figures in this thread), these were disappointing, reach is down 5.9% at 2052k, and total hours are down 11.8% at 11951k. Verdict: late summer/early autumn is probably never prime R3 listening time, but the Proms didn't seem to make much of an impact. CFM also had a similarly disappointing Q3, with a 5388k reach, down 6.4%, and 34135k total hours, down 12.3%. (Percentage changes are relative to previous quarter.)

R3 2011 Q4 results, released today, are an improvement compared to Q3, with reach up 2.1% at 2097k and total hours up an impressive 15.1% at 13762k. Overall R3 2011 figures are 2154k average reach, an increase of 4% over 2010, and total hours at 52866k, an increase of 11.6% over 2010. Overall verdict: R3 continues to gain audience*, but listeners are listening more.

Russ

* In context, this is a short-term year-to-year measure, i.e. the reach does not hit the heights of 1998 or 2003, but at least is far better than the nadir of 2007. The reach increase is probably not much compared to the total population though, as will probably be seen when johnb updates his marvellous spreadsheet graph.