Appreciation Index: Difference between revisions

From Tardis Wiki, the free Doctor Who reference
 
(33 intermediate revisions by 13 users not shown)
Line 1: Line 1:
{{real world}}
{{real world}}
The '''AI''' or '''Appreciation Index''' is a statistical representation of the amount of enjoyment the audience derived from a particular episode of television. Since ''[[Rose (TV story)|Rose]]'', it has been the method for measuring the tastes of the audience. Different methods were used during the 1963 version of the programme, and so BBC Wales AI scores are not compatible with audience research numbers given in the past, known until around [[1981 (production)|1981]] as the '''Reaction Index'''.  
The '''AI''' or '''Appreciation Index''' is a statistical representation of the amount of enjoyment the audience derived from a particular episode of television. Since ''[[Rose (TV story)|Rose]]'', it has been the method for measuring the tastes of the audience. Different methods were used during the 1963 version of the programme, and so BBC Wales AI scores are not compatible with audience research numbers given in the past, known until around 1981 as the '''Reaction Index'''.
 
== Modern measurement ==
== Modern measurement ==
A [[21st century]] AI score is calculated using a small but representative group of viewers. This sample will watch a program and then rate the program on a scale of one to ten. The scores are then averaged and multiplied by ten. Hence, an AI of 67 means that 6.7 was the simple mean of all responses. Scores of 85 or better are rare, and thus considered "excellent". Scores below 60 are considered "poor". Scores for every episode of the [[BBC Wales]] version of the program — save ''[[The End of the World (TV story)|The End of the World]]'' and ''[[Love & Monsters]]'' — have been 80 or above. Every story from ''[[Smith and Jones]]'' to ''[[Victory of the Daleks]]'' scored at least an 85 — a three-year run of the programme's most consistent and highest AI performances. The highest-ever AI for a ''Doctor Who'' episode was a 91, received by both ''[[The Stolen Earth]]'' and ''[[Journey's End]]''.
A 21st century AI score is calculated using a small but representative group of viewers. This sample will watch a program and then rate the program on a scale of one to ten. The scores are then averaged and multiplied by ten. Hence, an AI of 67 means that 6.7 was the simple mean of all responses. Scores of 85 or better are rare, and thus considered "excellent". Scores below 60 are considered "poor". Scores for every episode of the [[BBC Wales]] version of the program — {{as of|2019|1|alt=save}} ''[[The End of the World (TV story)|The End of the World]]'', ''[[Love & Monsters (TV story)|Love & Monsters]]'' and ''[[Sleep No More (TV story)|Sleep No More]];'' in [[series 11 (Doctor Who 2005)|series 11]]: ''[[The Tsuranga Conundrum (TV story)|The Tsuranga Conundrum]]'' and ''[[The Battle of Ranskoor Av Kolos (TV story)|The Battle of Ranskoor Av Kolos]]''; and in [[Series 12 (Doctor Who 2005)|series 12]]: ''[[Orphan 55 (TV story)|Orphan 55]]'', ''[[Nikola Tesla's Night of Terror (TV story)|Nikola Tesla's Night of Terror]]'', ''[[Praxeus (TV story)|Praxeus]]'', ''[[Can You Hear Me? (TV story)|Can You Hear Me?]]'' and ''[[Revolution of the Daleks (TV story)|Revolution of the Daleks]]'' and in [[Series 13 (Doctor Who 2005)|series 13]]: ''[[The Halloween Apocalypse (TV story)|The Halloween Apocalypse]]''— have been 80 or above. Every story from ''[[Smith and Jones (TV story)|Smith and Jones]]'' to ''[[Victory of the Daleks (TV story)|Victory of the Daleks]]'' scored at least an 85 — a three-year run of the programme's most consistent and highest AI performances. The highest-ever AI for a ''Doctor Who'' episode was a 91, received by both ''[[The Stolen Earth (TV story)|The Stolen Earth]]'' and ''[[Journey's End (TV story)|Journey's End]]''.


== History ==
== History ==
Determining "appreciation" was the original method for registering the success or failure of radio and television programs in [[Britain]]. It has been employed by the [[BBC]], in some form, since [[1936]]. With the advent of commercial television, however, advertisers wanted to know ''how many'' people were watching. Ratings — a measure of the raw number of viewers in front of television sets — supplanted the AI as the primary measure of televised success. However, as the BBC is not a commercial enterprise, the AI still retains importance in determining the fate of television programs on the networks of the BBC.  
Determining "appreciation" was the original method for registering the success or failure of radio and television programs in [[Britain]]. It has been employed by the [[BBC]], in some form, since 1936. With the advent of commercial television, however, advertisers wanted to know ''how many'' people were watching. Ratings — a measure of the raw number of viewers in front of television sets — supplanted the AI as the primary measure of televised success. However, as the BBC is not a commercial enterprise, the AI still retains importance in determining the fate of television programs on the networks of the BBC.
 
== The Reaction Index ==
== The Reaction Index ==
Audience research was done differently in the past. Until [[1981 (production)|1981]], the BBC performed their own internal research. It was carried out by the Audience Research Department, a fairly large branch of the BBC that had numerous field workers who actually went out into the country to personally question viewers and/or collect written surveys.<ref name=R9>[http://www.nationalarchives.gov.uk/nra/onlinelists/GB%200898%20R9.pdf "Series Level Description. Audience Research." R9/1-2,182. BBC Written Archives Centre.]</ref>
Audience research was done differently in the past. Until [[1981 (production)|1981]], the BBC performed their own internal research. It was carried out by the Audience Research Department, a fairly large branch of the BBC that had numerous field workers who actually went out into the country to personally question viewers and/or collect written surveys.<ref name=R9>[http://www.nationalarchives.gov.uk/nra/onlinelists/GB%200898%20R9.pdf "Series Level Description. Audience Research." R9/1-2,182. BBC Written Archives Centre.]</ref>


At least initially, reactions were measured with small samples of around 1000 people. [[An Unearthly Child (episode)|The first episode]]'s RI was calculated from a tiny 124 sample size, because only 12% of the ''full'' sample of about 1030 — called a "Viewing Panel" in internal documents — watched ''Doctor Who'' on [[23 November (releases)|23 November]] [[1963 (releases)|1963]]. They gave it a mark from a five-point scale, as follows:
At least initially, reactions were measured with small samples of around 1000 people. [[An Unearthly Child (episode)|The first episode]]'s RI was calculated from a tiny 124 sample size, because only 12% of the ''full'' sample of about 1030 — called a "Viewing Panel" in internal documents — watched ''Doctor Who'' on [[23 November (releases)|23 November]] [[1963 (releases)|1963]]. They gave it a mark from a five-point scale, as follows:
{|class=wikitable
{|class=wikitable
!A+
!A+
Line 21: Line 23:
This then allowed the calculation of an RI of 63, which was between the then-current averages of 62 for drama and 64 for children's programmes.<ref>[http://www.bbc.co.uk/archive/doctorwho/6406.shtml "Audience Research Report on 'An Unearthly Child'". Vr/63/668. BBC Archive.]</ref>
This then allowed the calculation of an RI of 63, which was between the then-current averages of 62 for drama and 64 for children's programmes.<ref>[http://www.bbc.co.uk/archive/doctorwho/6406.shtml "Audience Research Report on 'An Unearthly Child'". Vr/63/668. BBC Archive.]</ref>


== The first BARB AI ==
== BARB era ==
For the balance of the so-called "classic series", it was moved to a non-profit organisation called the {{W|Broadcasters' Audience Research Board}} or BARB, which was funded by all broadcasters.<ref name=R9/> During this time, respondents were asked to record a number between 0 and 6 for each programme they watched, with the following meanings<ref name=MRM>Gunter, Barrie. ''Media Research Methods: Measuring Audiences, Reaction and Impact.'' SAGE. 2000.</ref>:
For the balance of the so-called "classic series", it was moved to a non-profit organisation called the [[Broadcasters' Audience Research Board]] or BARB, which was funded by all broadcasters.<ref name=R9 /> During this time, respondents were asked to record a number between 0 and 6 for each programme they watched, with the following meanings<ref name=MRM>Gunter, Barrie. ''Media Research Methods: Measuring Audiences, Reaction and Impact.'' SAGE. 2000.</ref>:
{|class=wikitable
{|class=wikitable
!Respondent score
!Respondent score
Line 42: Line 44:
Thus, during most of the [[John Nathan-Turner]] era, an AI of 65 — a typical score — would have meant "only a little better than ''fairly interesting''".
Thus, during most of the [[John Nathan-Turner]] era, an AI of 65 — a typical score — would have meant "only a little better than ''fairly interesting''".


During the [[1990s]], a number of different scoring systems were tried, but they didn't result in significantly different AI results.<ref name=MRM/> Eventually, BARB settled on the simple ten point scale that has been in use throughout the [[BBC Wales]] era, and which has applied to ''[[Torchwood (TV series)|Torchwood]]'' and ''[[The Sarah Jane Adventures]]'' alike.
During the 1990s, a number of different scoring systems were tried, but they didn't result in significantly different AI results.<ref name=MRM /> Eventually, BARB settled on the simple ten point scale that has been in use throughout the [[BBC Wales]] era, and which has applied to ''[[Torchwood (TV series)|Torchwood]]'' and ''[[The Sarah Jane Adventures]]'' alike.
 
BARB continued calculating AIs as late as 2012. For most episodes of the BBC Wales era until 2012, the AI reported is that produced by BARB. However, since 2005, the BBC also received a second, separately-calculated AI score from {{w|GfK NOP}}, a separate organisation that used online surveys. When BARB bowed out, GFK, and its online "Pulse" panel took over entirely.


BARB continued calculating AIs as late as 2012. For most episodes of the BBC Wales era until 2012, the AI reported is that produced by BARB.  However, since 2005, the BBC also received a second, separately-calculated AI score from {{w|GfK NOP}}, a separate organisation that used online surveys.  When BARB bowed out, GFK, and its online "Pulse" panel took over entirely. 
== GfK era ==
== GfK Era ==
Since 2012, ''Doctor Who'' AIs have been exclusive collected from a large online sample of some 20,000 individuals responding to an online survey. According to this measurement, 81 is the average [[BBC One]] AI. The basic response system is the same ten point scale as BARB used, meaning that an AI of 81 corresponds to an average respondent score of 8.1.
Since 2012, ''Doctor Who'' AIs have been exclusive collected from a large online sample of some 20,000 individuals responding to an online survey. Accroding to this measurement, 81 is the average [[BBC One]] AI. The basic response system is the same ten point scale as BARB used, meaning that an AI of 81 corresponds to an average respondent score of 8.1.


The online nature of the survey necessarily excludes those without internet access from participating, possibly marginalising the voices of the elderly, the poor, and the technophobes. However, it also allows for more sophisticated weighting of respondents. According to a BBC report, "The daily reporting panel is weighted for age, social grade, sex, presence of children, region and the household digital type on a daily basis to ensure it is representative of the UK as a whole."<ref>[http://www.bbc.co.uk/archive/doctorwho/6406.shtml "BBC Audience Information: January - March 2011". BBC Head of Audiences.]</ref>
The online nature of the survey necessarily excludes those without internet access from participating, possibly marginalising the voices of the elderly, the poor, and the technophobes. However, it also allows for more sophisticated weighting of respondents. According to a BBC report, "The daily reporting panel is weighted for age, social grade, sex, presence of children, region and the household digital type on a daily basis to ensure it is representative of the UK as a whole."<ref>[http://www.bbc.co.uk/archive/doctorwho/6406.shtml "BBC Audience Information: January - March 2011". BBC Head of Audiences.]</ref>


== Footnotes ==  
== Footnotes ==
{{reflist}}  
{{reflist}}
{{wikipediainfo}}
 
[[Category:Terminology]]
[[Category:Terminology]]

Latest revision as of 20:51, 25 April 2024

RealWorld.png

The AI or Appreciation Index is a statistical representation of the amount of enjoyment the audience derived from a particular episode of television. Since Rose, it has been the method for measuring the tastes of the audience. Different methods were used during the 1963 version of the programme, and so BBC Wales AI scores are not compatible with audience research numbers given in the past, known until around 1981 as the Reaction Index.

Modern measurement[[edit] | [edit source]]

A 21st century AI score is calculated using a small but representative group of viewers. This sample will watch a program and then rate the program on a scale of one to ten. The scores are then averaged and multiplied by ten. Hence, an AI of 67 means that 6.7 was the simple mean of all responses. Scores of 85 or better are rare, and thus considered "excellent". Scores below 60 are considered "poor". Scores for every episode of the BBC Wales version of the program — save The End of the World, Love & Monsters and Sleep No More; in series 11: The Tsuranga Conundrum and The Battle of Ranskoor Av Kolos; and in series 12: Orphan 55, Nikola Tesla's Night of Terror, Praxeus, Can You Hear Me? and Revolution of the Daleks and in series 13: The Halloween Apocalypse— have been 80 or above. Every story from Smith and Jones to Victory of the Daleks scored at least an 85 — a three-year run of the programme's most consistent and highest AI performances. The highest-ever AI for a Doctor Who episode was a 91, received by both The Stolen Earth and Journey's End.

History[[edit] | [edit source]]

Determining "appreciation" was the original method for registering the success or failure of radio and television programs in Britain. It has been employed by the BBC, in some form, since 1936. With the advent of commercial television, however, advertisers wanted to know how many people were watching. Ratings — a measure of the raw number of viewers in front of television sets — supplanted the AI as the primary measure of televised success. However, as the BBC is not a commercial enterprise, the AI still retains importance in determining the fate of television programs on the networks of the BBC.

The Reaction Index[[edit] | [edit source]]

Audience research was done differently in the past. Until 1981, the BBC performed their own internal research. It was carried out by the Audience Research Department, a fairly large branch of the BBC that had numerous field workers who actually went out into the country to personally question viewers and/or collect written surveys.[1]

At least initially, reactions were measured with small samples of around 1000 people. The first episode's RI was calculated from a tiny 124 sample size, because only 12% of the full sample of about 1030 — called a "Viewing Panel" in internal documents — watched Doctor Who on 23 November 1963. They gave it a mark from a five-point scale, as follows:

A+ A B C C-
17% 39% 31% 6% 7%

This then allowed the calculation of an RI of 63, which was between the then-current averages of 62 for drama and 64 for children's programmes.[2]

BARB era[[edit] | [edit source]]

For the balance of the so-called "classic series", it was moved to a non-profit organisation called the Broadcasters' Audience Research Board or BARB, which was funded by all broadcasters.[1] During this time, respondents were asked to record a number between 0 and 6 for each programme they watched, with the following meanings[3]:

Respondent score BARB translation What it meant
1 0 not at all interesting and/or enjoyable
2 20 not very interesting and/or enjoyable
3 40 neither one thing nor the other
4 60 fairly interesting and/or enjoyable
5 80 very interesting and/or enjoyable
6 100 extremely interesting and/or enjoyable

Thus, during most of the John Nathan-Turner era, an AI of 65 — a typical score — would have meant "only a little better than fairly interesting".

During the 1990s, a number of different scoring systems were tried, but they didn't result in significantly different AI results.[3] Eventually, BARB settled on the simple ten point scale that has been in use throughout the BBC Wales era, and which has applied to Torchwood and The Sarah Jane Adventures alike.

BARB continued calculating AIs as late as 2012. For most episodes of the BBC Wales era until 2012, the AI reported is that produced by BARB. However, since 2005, the BBC also received a second, separately-calculated AI score from GfK NOP, a separate organisation that used online surveys. When BARB bowed out, GFK, and its online "Pulse" panel took over entirely.

GfK era[[edit] | [edit source]]

Since 2012, Doctor Who AIs have been exclusive collected from a large online sample of some 20,000 individuals responding to an online survey. According to this measurement, 81 is the average BBC One AI. The basic response system is the same ten point scale as BARB used, meaning that an AI of 81 corresponds to an average respondent score of 8.1.

The online nature of the survey necessarily excludes those without internet access from participating, possibly marginalising the voices of the elderly, the poor, and the technophobes. However, it also allows for more sophisticated weighting of respondents. According to a BBC report, "The daily reporting panel is weighted for age, social grade, sex, presence of children, region and the household digital type on a daily basis to ensure it is representative of the UK as a whole."[4]

Footnotes[[edit] | [edit source]]