1. Robert Ford says: • Website

   October 6, 2013 at 3:40 pm GMT • 100 Words

   Z, are we gonna get a post expanding on what you learned about Lewontin’s fallacy applying to dogs or did you already do one that covers that?

   http://www.youtube.com/watch?v=NxLkEPb5DzI
   Also, I’d like to encourage everyone to soak in this Charles Barkley golf swing montage – at least watch until about 1:10 when he hits someone in the neck:)

   • Replies: @Razib Khan

   @Robert Ford

   the fallacy is the same. turns out most variation in dog breeds is *within* groups, there was just ascertainment bias because of original focus on genes which have a high Fst between groups. the differences between breeds probably has more to do with moderate differences in allele frequencies across many genes, instead of big ones on a few.

2. T. Greer says: • Website

   October 6, 2013 at 7:02 pm GMT • 200 Words

   I have one. What is up with this map?

   The map if of sexual violence in India. And I have trouble making sense of it. As I say in that post:

   “I was rather astonished by this data. I expected sexual violence to correspond with development levels. This is not the case. Sexual violence in Assam, West Bengal, and Rajasthan, states with low rates of GDP per capita, literacy, life expectancy, and other common indexes of development, is barely worse than in Kerala, a state that leads India in almost all of these measures. Strange combinations can also be found at the bottom of the list, where rich developed states like Goa have the same rate of sexual violence as Bihar, India’s poorest and least educated state.

   Neither wealth, knowledge, or living standards can explain the distribution of India’s sexual violence. What does? I have spent the last thirty minutes looking at lists of India’s states by sex ratio, household size, HIV awareness, corruption, language, family structure, and number of Muslims in major metropolitan areas. None of these lists match our pattern. There does seem to be a relation between sexual violence and violent crimes generally but that does not explain too much. We are still left asking why some places in India are plagued with violence while others are relatively free of this problem.

   Razib, do you see any patterns in the data that I am missing?

   http://www.scholars-stage.blogspot.com/2013/10/trying-to-make-sense-of-indias-sexual.html

   • Replies: @Razib Khan

   @T. Greer

   nope. how much can we trust them though?

   , @Anthony_A

   @T. Greer

   I'll hazard a guess: Sexual violence is lower among wealthier/more educated populations, but is also more likely to be reported in those populations. With reporting rates correlated to incidence rates, what you're seeing is the residual between those correlations, and thus seemingly random patterns, or secondary effects on one or the other rate appearing much more important than they really are.

   For example, if Muslims were somewhat less (or more) likely to report sexual violence than otherwise-demographically-similar Hindus, but no more or less likely to commit such violence, the reported rates would seem lower (or higher) by Muslim population concentration, even though the actual incidence was purely a function of income and education, and not religion.

3. Razib Khan says:

   October 7, 2013 at 12:40 am GMT • 100 Words

   @Robert Ford

   Z, are we gonna get a post expanding on what you learned about Lewontin's fallacy applying to dogs or did you already do one that covers that?

   http://www.youtube.com/watch?v=NxLkEPb5DzI
   Also, I'd like to encourage everyone to soak in this Charles Barkley golf swing montage - at least watch until about 1:10 when he hits someone in the neck:)

   Replies: @Razib Khan

   the fallacy is the same. turns out most variation in dog breeds is *within* groups, there was just ascertainment bias because of original focus on genes which have a high Fst between groups. the differences between breeds probably has more to do with moderate differences in allele frequencies across many genes, instead of big ones on a few.
4. Razib Khan says:

   October 7, 2013 at 12:41 am GMT

   @T. Greer

   I have one. What is up with this map?

   The map if of sexual violence in India. And I have trouble making sense of it. As I say in that post:

   
   "I was rather astonished by this data. I expected sexual violence to correspond with development levels. This is not the case. Sexual violence in Assam, West Bengal, and Rajasthan, states with low rates of GDP per capita, literacy, life expectancy, and other common indexes of development, is barely worse than in Kerala, a state that leads India in almost all of these measures. Strange combinations can also be found at the bottom of the list, where rich developed states like Goa have the same rate of sexual violence as Bihar, India's poorest and least educated state.

   Neither wealth, knowledge, or living standards can explain the distribution of India's sexual violence. What does? I have spent the last thirty minutes looking at lists of India's states by sex ratio, household size, HIV awareness, corruption, language, family structure, and number of Muslims in major metropolitan areas. None of these lists match our pattern. There does seem to be a relation between sexual violence and violent crimes generally but that does not explain too much. We are still left asking why some places in India are plagued with violence while others are relatively free of this problem.
    

   Razib, do you see any patterns in the data that I am missing?

   
   http://www.scholars-stage.blogspot.com/2013/10/trying-to-make-sense-of-indias-sexual.html

   Replies: @Razib Khan, @Anthony_A

   nope. how much can we trust them though?
5. Anthony_A says:

   October 7, 2013 at 8:05 pm GMT • 100 Words

   @T. Greer

   I have one. What is up with this map?

   The map if of sexual violence in India. And I have trouble making sense of it. As I say in that post:

   
   "I was rather astonished by this data. I expected sexual violence to correspond with development levels. This is not the case. Sexual violence in Assam, West Bengal, and Rajasthan, states with low rates of GDP per capita, literacy, life expectancy, and other common indexes of development, is barely worse than in Kerala, a state that leads India in almost all of these measures. Strange combinations can also be found at the bottom of the list, where rich developed states like Goa have the same rate of sexual violence as Bihar, India's poorest and least educated state.

   Neither wealth, knowledge, or living standards can explain the distribution of India's sexual violence. What does? I have spent the last thirty minutes looking at lists of India's states by sex ratio, household size, HIV awareness, corruption, language, family structure, and number of Muslims in major metropolitan areas. None of these lists match our pattern. There does seem to be a relation between sexual violence and violent crimes generally but that does not explain too much. We are still left asking why some places in India are plagued with violence while others are relatively free of this problem.
    

   Razib, do you see any patterns in the data that I am missing?

   
   http://www.scholars-stage.blogspot.com/2013/10/trying-to-make-sense-of-indias-sexual.html

   Replies: @Razib Khan, @Anthony_A

   I’ll hazard a guess: Sexual violence is lower among wealthier/more educated populations, but is also more likely to be reported in those populations. With reporting rates correlated to incidence rates, what you’re seeing is the residual between those correlations, and thus seemingly random patterns, or secondary effects on one or the other rate appearing much more important than they really are.

   For example, if Muslims were somewhat less (or more) likely to report sexual violence than otherwise-demographically-similar Hindus, but no more or less likely to commit such violence, the reported rates would seem lower (or higher) by Muslim population concentration, even though the actual incidence was purely a function of income and education, and not religion.
6. Joe Q. says:

   October 9, 2013 at 4:14 pm GMT • 100 Words

   Razib — yet another paper recently appeared on the genetics of the Ashkenazi Jews.

   http://www.nature.com/ncomms/2013/131008/ncomms3543/full/ncomms3543.html

   I am interested in the topic but am nowhere near qualified to understand the details (beyond the introduction). My take on it is that (1) it corroborates the idea that the Ashkenazim are, very broadly, descendants of Near Eastern (“Judaean”) men who lived in the Mediterranean basin and took local (“Roman”) wives, and that (2) it puts another nail in the coffin of the “Khazar hypothesis”.

   Am I missing something important?

   • Replies: @Razib Khan

   @Joe Q.

   probably. i want to look at X chromosomes though. that would probably clarify and clean up, as uniparental assignments can be dicey.

7. Razib Khan says:

   October 9, 2013 at 6:29 pm GMT

   @Joe Q.

   Razib -- yet another paper recently appeared on the genetics of the Ashkenazi Jews.

   http://www.nature.com/ncomms/2013/131008/ncomms3543/full/ncomms3543.html

   I am interested in the topic but am nowhere near qualified to understand the details (beyond the introduction). My take on it is that (1) it corroborates the idea that the Ashkenazim are, very broadly, descendants of Near Eastern ("Judaean") men who lived in the Mediterranean basin and took local ("Roman") wives, and that (2) it puts another nail in the coffin of the "Khazar hypothesis".

   
   Am I missing something important?

   Replies: @Razib Khan

   probably. i want to look at X chromosomes though. that would probably clarify and clean up, as uniparental assignments can be dicey.

Comments are closed.

Subscribe to All Razib Khan Comments via RSS