CCBC-Net Archives
STEM and Representation follow-up
- Contemporary messages sorted: [ by date ] [ by subject ] [ by author ]
From: Charles Bayless <charles.bayless_at_ttmd.com>
Date: Fri, 29 Mar 2013 14:14:58 -0400
I have waited till the end of the month in order to not distract from the current discussion. I have a follow-up to last month's conversation regarding representation in children's literature. Recapping: The argument was advanced 1) that certain classes of children (race, gender, class, culture, orientation, etc.) are underrepresented in children's literature, 2) that this lack of representation creates a material negative impact on children, and 3) that the cause of such underrepresentation was conscious or unconscious discrimination on the part of the publishing industry.
The counterargument was that a) there is no objective empirical evidence that documents the extent and degree of underrepresentation by group (not contesting that there likely was underrepresentation, only that there was no data to quantify it), b) that there is no empirical evidence supporting any impact of over or underrepresentation of traditional categorizations (race, gender, class, orientation, etc.) on life outcomes and much obvious evidence to the contrary, and c) that there was no evidence supporting discrimination on the part of the publishing industry.
In the spirit of March's STEM theme (science being the generation and testing of hypotheses with data) I have been able to source data from credible organizations (US Census, US Bureau of Labor Statistics, CCBC and the Kaiser Foundation) that allows a first order crack at answering argument element 1 (degree of representation). In doing so, that also begins to address argument element 3 (publishing industry discrimination). In order not to disrupt the CCBC monthly conversation, I have posted the methodology, details of analysis, and results at my blog, Thing Finder - Reading, Publishing, and Ethnicity ml). Comments are welcome there.
This analysis is only a first order estimate. It is always better to use data collected in a common controlled fashion than to use data from multiple sources. In the absence of any other evidence though, this is likely a reasonably robust first order effort.
The summary is that at a demographic level, all minority groups are underrepresented in children's books. However, publishers being commercial enterprises, they are not responding to the raw census level of population but rather to the subpopulation that are book buyers. What I have done is take the average reading volume (hours per year) for each group, adjusted it for the percentage of reading that is book related (rather than magazines and newspapers) and adjusted yet further to reflect propensity to purchase books. This allows us to assess whether and to what degree publishers are responding to the book buying public. The answer is that the non-Hispanic white book buying public is being served proportionately, that the African-American and Native American book buying populations are being dramatically overrepresented and the Hispanic and Asian American populations are being materially underrepresented.
The gap between demographics and the book buying populations arises because of differentials in the volume of book reading and book buying being done by the respective groups. The necessary corollary conclusion is that increased supply won't do much if there is not increased demand. (Also see Richard Nash's article What is the business of Literature?
in which he observes that "Abundance, it turns out, is a much bigger problem to solve than scarcity").
The fact that publishers are significantly oversupplying some groups and undersupplying others would seem to clear publishers of the charge of conscious or unconscious discrimination. It suggests that they are possibly ineffective at correctly judging the extent to which different groups would be willing to buy themed books but the data indicates that they are clearly attempting to address submarkets.
So we now have a first order estimate of the degree of over and underrepresentation by ethnicity and we also have data indicating that simple discrimination within the publishing industry is unlikely the root cause of lack of representation for some groups. Thought you might be interested in the analysis.
None of this addresses the core issue, whether there is any data to support the contention that representation has any measurable impact on life outcomes. If there is no measurable impact of over or underrepresentation, then the whole issue is moot and something of a red herring.
While we are on a data analysis roll, I have another post which might be of interest to teachers and librarians. The US education system frequently gets criticized in international comparisons as both expensive and ineffective, with our students scoring only in the middle of the pack. In the post at Thing Finder - U.S. Education Expensive and Ineffective? Not So Fast ive.html), I look at the underlying data. The US system is actually extremely effective (at least in terms of the reading aspect of education) when one compares like-to-like in terms of cultural origins, scoring better than all sixty participants save only Finland.
Charles
Received on Fri 29 Mar 2013 02:14:58 PM CDT
Date: Fri, 29 Mar 2013 14:14:58 -0400
I have waited till the end of the month in order to not distract from the current discussion. I have a follow-up to last month's conversation regarding representation in children's literature. Recapping: The argument was advanced 1) that certain classes of children (race, gender, class, culture, orientation, etc.) are underrepresented in children's literature, 2) that this lack of representation creates a material negative impact on children, and 3) that the cause of such underrepresentation was conscious or unconscious discrimination on the part of the publishing industry.
The counterargument was that a) there is no objective empirical evidence that documents the extent and degree of underrepresentation by group (not contesting that there likely was underrepresentation, only that there was no data to quantify it), b) that there is no empirical evidence supporting any impact of over or underrepresentation of traditional categorizations (race, gender, class, orientation, etc.) on life outcomes and much obvious evidence to the contrary, and c) that there was no evidence supporting discrimination on the part of the publishing industry.
In the spirit of March's STEM theme (science being the generation and testing of hypotheses with data) I have been able to source data from credible organizations (US Census, US Bureau of Labor Statistics, CCBC and the Kaiser Foundation) that allows a first order crack at answering argument element 1 (degree of representation). In doing so, that also begins to address argument element 3 (publishing industry discrimination). In order not to disrupt the CCBC monthly conversation, I have posted the methodology, details of analysis, and results at my blog, Thing Finder - Reading, Publishing, and Ethnicity ml). Comments are welcome there.
This analysis is only a first order estimate. It is always better to use data collected in a common controlled fashion than to use data from multiple sources. In the absence of any other evidence though, this is likely a reasonably robust first order effort.
The summary is that at a demographic level, all minority groups are underrepresented in children's books. However, publishers being commercial enterprises, they are not responding to the raw census level of population but rather to the subpopulation that are book buyers. What I have done is take the average reading volume (hours per year) for each group, adjusted it for the percentage of reading that is book related (rather than magazines and newspapers) and adjusted yet further to reflect propensity to purchase books. This allows us to assess whether and to what degree publishers are responding to the book buying public. The answer is that the non-Hispanic white book buying public is being served proportionately, that the African-American and Native American book buying populations are being dramatically overrepresented and the Hispanic and Asian American populations are being materially underrepresented.
The gap between demographics and the book buying populations arises because of differentials in the volume of book reading and book buying being done by the respective groups. The necessary corollary conclusion is that increased supply won't do much if there is not increased demand. (Also see Richard Nash's article What is the business of Literature?
in which he observes that "Abundance, it turns out, is a much bigger problem to solve than scarcity").
The fact that publishers are significantly oversupplying some groups and undersupplying others would seem to clear publishers of the charge of conscious or unconscious discrimination. It suggests that they are possibly ineffective at correctly judging the extent to which different groups would be willing to buy themed books but the data indicates that they are clearly attempting to address submarkets.
So we now have a first order estimate of the degree of over and underrepresentation by ethnicity and we also have data indicating that simple discrimination within the publishing industry is unlikely the root cause of lack of representation for some groups. Thought you might be interested in the analysis.
None of this addresses the core issue, whether there is any data to support the contention that representation has any measurable impact on life outcomes. If there is no measurable impact of over or underrepresentation, then the whole issue is moot and something of a red herring.
While we are on a data analysis roll, I have another post which might be of interest to teachers and librarians. The US education system frequently gets criticized in international comparisons as both expensive and ineffective, with our students scoring only in the middle of the pack. In the post at Thing Finder - U.S. Education Expensive and Ineffective? Not So Fast ive.html), I look at the underlying data. The US system is actually extremely effective (at least in terms of the reading aspect of education) when one compares like-to-like in terms of cultural origins, scoring better than all sixty participants save only Finland.
Charles
Received on Fri 29 Mar 2013 02:14:58 PM CDT
