Methods and apparatus to assign viewers to media meter data
View Patent ↗Methods, apparatus, systems and articles of manufacture are disclosed to assign viewers to media meter data. An example method includes calculating, with a processor, first viewing probabilities for tuning panelists in a tuning household during a first set of time periods, calculating, with the processor, second viewing probabilities for viewing panelists in a plurality of candidate viewing households during a second set of time periods, identifying, with the processor, a matching one of the plurality of candidate viewing households based on an absolute difference value between an average value of the first viewing probabilities and respective ones of average values of the second viewing probabilities, and reducing an imputation error by imputing, with the processor, tuning minutes of the tuning household as viewing minutes for the respective tuning panelists when the matching one of the plurality of candidate viewing households exhibits viewing activity during one of the second set of time periods that matches one of the first set of time periods.
1. A method to impute panelist household viewing behavior, the method comprising:
monitoring, via a first meter, a first media presentation device in a tuning household associated with tuning panelists, the first meter structured to collect first media identification data indicative of first media to which the first media presentation device is tuned without collecting person identifying information indicative of which of the tuning panelists are exposed to the first media;
monitoring, via second meters, second media presentation devices in a set of viewing households associated with viewing panelists, the second meters structured to (i) collect second media identification data indicative of respective second media to which the second media presentation devices are tuned and (ii) collect person identifying information indicative of which of the viewing panelists are exposed to the respective second media;
transmitting, via a network, first data from the first meter, the first data including the first media identification data, the first data not including the person identifying information associated with the tuning panelists;
transmitting, via the network, second data from the second meters, the second data including the second media identification data and the person identifying information associated with the viewing panelists;
calculating, by executing an instruction with a processor, first viewing probabilities for the tuning panelists during a first set of time periods, the first viewing probabilities calculated based on the first data;
identifying a plurality of candidate viewing households from among the set of viewing households based on a similarity of household characteristics with the tuning household;
calculating, by executing an instruction with the processor, second viewing probabilities for the viewing panelists in the plurality of candidate viewing households during a second set of time periods, the second viewing probabilities calculated based on the second data;
identifying, by executing an instruction with the processor, a matching one of the plurality of candidate viewing households that matches the tuning household based on an absolute difference value between an average value of the first viewing probabilities and respective ones of average values of the second viewing probabilities; and
imputing, by executing an instruction with the processor, ones of tuning minutes of the tuning household as viewing minutes for ones of the tuning panelists when the matching one of the plurality of candidate viewing households exhibits viewing activity during one of the second set of time periods that matches one of the first set of time periods, the tuning minutes indicative of when the first media presentation device was tuned to the first media, the viewing minutes indicative of when the ones of the tuning panelists were exposed to the first media to which the first media presentation device was tuned.
2. The method as defined in claim 1 , wherein the first set of time periods and the second set of time periods include data points having a quarter hour resolution.
3. The method as defined in claim 1 , further including calculating an adjusted quarter hour ratio when a number of the first set of time periods is dissimilar to a number of the second set of time periods.
4. The method as defined in claim 3 , wherein the adjusted quarter hour ratio includes a ratio of the number of the second set of time periods divided by the number of the first set of time periods.
5. The method as defined in claim 4 , further including generating a final quarter hour order value associated with the second set of time periods by multiplying the adjusted quarter hour ratio by a temporal placeholder associated with the first set of time periods.
6. The method as defined in claim 5 , further including expanding the number of the second set of time periods to match the first set of time periods based on the final quarter hour order value.
7. The method as defined in claim 1 , wherein identifying the matching one of the plurality of candidate viewing households based on the absolute difference value further includes selecting the matching one of the plurality of candidate viewing households that do not exhibit viewing behavior within a similar geographic area when the absolute difference value satisfies a threshold value from a next-lowest absolute difference value.
8. A system to impute panelist household viewing behavior, the system comprising:
a first meter to monitor a first media presentation device in a tuning household associated with tuning panelists, the first meter structured to collect first media identification data indicative of first media to which the first media presentation device is tuned without collecting person identifying information indicative of which of the tuning panelists are exposed to the first media;
second meters to monitor second media presentation devices in a set of viewing households associated with viewing panelists, the second meters structured to (i) collect second media identification data indicative of respective second media to which the second media presentation devices are tuned and (ii) collect person identifying information indicative of which of the viewing panelists are exposed to the respective second media;
an independent distribution probability selector to identify a plurality of candidate viewing households from among the set of viewing households based on a similarity of household characteristics with a tuning household;
a total probability calculator to:
calculate first viewing probabilities for the tuning panelists during a first set of time periods, the first viewing probabilities calculated based on first data received, via a network, from the first meter, the first data including the first media identification data, the first data not including the person identifying information associated with the tuning panelists; and
calculate second viewing probabilities for the viewing panelists in the plurality of candidate viewing households during a second set of time periods, the second viewing probabilities calculated based on second data received, via the network, from the second meters, the second data including the second media identification data and the person identifying information associated with the viewing panelists; and
a rank engine to:
identify a matching one of the plurality of candidate viewing households that matches the tuning household based on an absolute difference value between an average value of the first viewing probabilities and respective ones of average values of the second viewing probabilities; and
impute ones of tuning minutes of the tuning household as viewing minutes for ones of the tuning panelists when the matching one of the plurality of candidate viewing households exhibits viewing activity during one of the second set of time periods that matches one of the first set of time periods, wherein at least one of the independent distribution probability selector, the total probability calculator, or the rank engine is implemented using a processor, the tuning minutes indicative of when the first media presentation device was tuned to the first media, the viewing minutes indicative of when the ones of the tuning panelists were exposed to the first media to which the first media presentation device was tuned.
9. The system as defined in claim 8 , wherein the first set of time periods and the second set of time periods include data points having a quarter hour resolution.
10. The system as defined in claim 8 , further including a minutes aggregator to calculate an adjusted quarter hour ratio when a number of the first set of time periods is dissimilar to a number of the second set of time periods.
11. The system as defined in claim 10 , wherein the minutes aggregator is to calculate the adjusted quarter hour ratio as a ratio of the number of the second set of time periods and the number of the first set of time periods.
12. The system as defined in claim 11 , wherein the minutes aggregator is to generate a final quarter hour order value associated with the second set of time periods by multiplying the adjusted quarter hour ratio by a temporal placeholder associated with the first set of time periods.
13. The system as defined in claim 12 , wherein the minutes aggregator is to expand the number of the second set of time periods to match the first set of time periods based on the final quarter hour order value.
14. The system as defined in claim 8 , wherein the rank engine is to select the matching one of the plurality of candidate viewing households that do not exhibit viewing behavior within a similar geographic area when the absolute difference value satisfies a threshold value from a next-lowest absolute difference value.
15. A non-transitory computer readable medium comprising computer readable instructions that, when executed, cause a processor to at least:
retrieve first data transmitted from a first meter monitoring a first media presentation device in a tuning household associated with tuning panelists, the first meter structured to collect first media identification data indicative of first media to which the first media presentation device was tuned without collecting person identifying information indicative of which of the tuning panelists were exposed to the first media, the first data including the first media identification data, the first data not including the person identifying information associated with the tuning panelists;
retrieve second data transmitted from a second meter monitoring second media presentation devices in a set of viewing households associated with viewing panelists, the second meters structured to (i) collect second media identification data indicative of respective second media to which the second media presentation devices were tuned and (ii) collect person identifying information indicative of which of the viewing panelists were exposed to the respective second media, the second data including the second media identification data and the person identifying information associated with the viewing panelists;
calculate first viewing probabilities for the tuning panelists during a first set of time periods, the first viewing probabilities calculated based on the first data;
identify a plurality of candidate viewing households from among the set of viewing households based on a similarity of household characteristics with the tuning household;
calculate second viewing probabilities for the viewing panelists in the plurality of candidate viewing households during a second set of time periods, the second viewing probabilities calculated based on the second data;
identify a matching one of the plurality of candidate viewing households that matches the tuning household based on an absolute difference value between an average value of the first viewing probabilities and respective ones of average values of the second viewing probabilities; and
impute ones of tuning minutes of the tuning household as viewing minutes for ones of the tuning panelists when the matching one of the plurality of candidate viewing households exhibits viewing activity during one of the second set of time periods that matches one of the first set of time periods, the tuning minutes indicative of when the first media presentation device was tuned to the first media, the viewing minutes indicative of when the ones of the tuning panelists were exposed to the first media to which the first media presentation device was tuned.
16. The non-transitory computer readable medium as defined in claim 15 , wherein the instructions, when executed, cause the processor to calculate an adjusted quarter hour ratio when a number of the first set of time periods is dissimilar to a number of the second set of time periods.
17. The non-transitory computer readable medium as defined in claim 16 , wherein the instructions, when executed, cause the processor to compute a ratio of the number of the second set of time periods and the number of the first set of time periods.
18. The non-transitory computer readable medium as defined in claim 17 , wherein the instructions, when executed, cause the processor to generate a final quarter hour order value associated with the second set of time periods by multiplying the adjusted quarter hour ratio by a temporal placeholder associated with the first set of time periods.
19. The non-transitory computer readable medium as defined in claim 18 , wherein the instructions, when executed, cause the processor to expand the number of the second set of time periods to match the first set of time periods based on the final quarter hour order value.
20. The non-transitory computer readable medium as defined in claim 15 , wherein the instructions, when executed, cause the processor to select the matching one of the plurality of candidate viewing households that do not exhibit viewing behavior within a similar geographic area when the absolute difference value satisfies a threshold value from a next-lowest absolute difference value.