Menu
Jump to content

What's with the ads?

AFwife Claire

Fascinating (But Sad) Wired Article on Poverty and Child Abuse Prediction

Recommended Posts

I have been reading the threads on child poverty with deep interest and sorrow at what some members here have gone through.  Today I happened to read an article in Wired magazine called "A Child Abuse Prediction Model Fails Poor Families".  

 

Basically, Allegheny County, PA is using this model to predict child abuse, but the model uses data mined from public service records, such as child welfare, Head Start, county housing authority, drug and alcohol services-pretty much all services that only poor people use--in essence, reporting "how many public resources families have consumed", and using that to predict child abuse risk.

 

Here was the paragraph that really struck a chord in me:  

 

"We might call this poverty profiling. Like racial profiling, poverty profiling targets individuals for extra scrutiny based not on their behavior but rather on a personal characteristic: They live in poverty. Because the model confuses parenting while poor with poor parenting, the AFST views parents who reach out to public programs as risks to their children."

 

Anyway, this just seems like one more thing designed to help people that in actuality bites them in the rear.  The author of the article suggested averaging the model's score with one a human worker comes up with, but that doesn't happen, and it doesn't seem like it ever will.  Machines are always right, and certainly not prejudiced, of course!

 

 

  • Like 4

Share this post


Link to post
Share on other sites

Nailed it - parenting while poor does NOT equal poor parenting, and models based on that will fail because they are fundamentally flawed in their assumptions of correlation vs causation (ie That some causes of poverty, like substance abuse or mental illness, are causative of maladaptive and abusive parenting behaviors. But not the poverty, itself).

 

And this needs to be in sky writing and neon lights, and on way more subjects than just this:

 

But, as data scientist Cathy O’Neil has written, “models are opinions embedded in mathematics.â€

Edited by Arctic Mama
  • Like 11

Share this post


Link to post
Share on other sites

So the children of parents who reach out for needed help are judged more at risk than parents that say... only reach out to Elvis impersonators for renewal of marriage vows? 

  • Like 10

Share this post


Link to post
Share on other sites

https://www.nytimes.com/2018/01/02/magazine/can-an-algorithm-tell-when-kids-are-in-danger.html?mc=aud_dev&mcid=fb-nytimes&mccr=JanMC34&mcdt=2018-01&subid=JanMC34&ad-keywords=AudDevGate

 

I saw a New York Times article recently about that from a different perspective. I will have to read the one you linked later.

  • Like 2

Share this post


Link to post
Share on other sites

It also shows a complete lack of faith in their intervention programmes. They should after all be reducing the risk.

  • Like 2

Share this post


Link to post
Share on other sites

Yikes. This is why correlation science can often times run amok. Looking at large swaths of statistical data one can draw correlations with many things.

 

The real issue with why poverty precipitates out in this way, unfortunately, is because you will find a higher number of people who are poor and who abuse. Often this isn't due to the poverty of course, it is due to poor self control (issues with executive functioning, impulsivity, reactivity etc) which might lead to both being poor and being abusive. Much like the propensity to abuse, being poor is a factor of that struggle with self control and not due to being poor in and of itself. So basically, more people who abuse will be poor because being reactive and impulsive or having low grit and perserverence does not often lead to high educational attainment or the steps necessary to love into a higher income bracket. However, this does not work the other way around (people who are poor are not more likely to abuse) because being poor isn't always due to an inability to delay gratification long enough to get ahead. Sometimes it is a choice to choose a low paying career, sometimes it is health related, sometimes it is sacrifice to be with the children...this is not a situation that often leads to abuse. Trying to take human behavior and distilling it down into a mathematical function is just not helpful at all. This is exactly how biases and prejudices begin to precipitate within a culture. It is gross and strips people of their dignity.

  • Like 2

Share this post


Link to post
Share on other sites

It's yet another no-win situation.   

 

We can say that not every child in poverty is abused and not every abused child is in poverty and be correct, but there is a known relationship, even though it isn't 1:1.  If we target these families for closer scrutiny, we mess with innocent families and probably create an extra barrier to help for *all families.  But, if we ignore the relationship that does exist, aren't we knowingly making it harder to reduce abuse?

 

The conversations going on right now weigh heavily on me... as a homeschooler, as someone who's been investigated (for non-homeschool-related false accusations), as someone with young relatives who are being failed by adults, in both the family and the system, as someone with other relatives who work in the system, as someone who lives in a high poverty area with many serious issues, and as someone who's lived at or near poverty level in the past with zero abuse.

 

I'm not doing anything wrong and I don't want to be messed with. That's true.  But SOMETHING needs to be done.

 

Like the whole school reform thing, if we sit around waiting for the perfect solution, we're going to lose kids.

  • Like 2

Share this post


Link to post
Share on other sites

It's yet another no-win situation.   

 

We can say that not every child in poverty is abused and not every abused child is in poverty and be correct, but there is a known relationship, even though it isn't 1:1.  If we target these families for closer scrutiny, we mess with innocent families and probably create an extra barrier to help for *all families.  But, if we ignore the relationship that does exist, aren't we knowingly making it harder to reduce abuse?

 

Probably, but on the other hand, we're reinforcing the class privilege of middle and upper class abusers. They're respectable, so they're given a pass on things poorer people don't get a pass on. btdt

Share this post


Link to post
Share on other sites

Probably, but on the other hand, we're reinforcing the class privilege of middle and upper class abusers. They're respectable, so they're given a pass on things poorer people don't get a pass on. btdt

 

I absolutely agree that it would be an unintentional, very negative consequence. 

 

I'm sure I have a lot of personal bias, as I think it could potentially save my own niece and nephew.  On the other hand, I'm acutely aware of the eyes our social worker had on our tiny little house with our messy yard when he popped over and we were knocking around in our knocking around the tiny little house and messy yard clothes. I don't recall any income questions.  Any observer would likely assume it to be low, at best.

  • Like 1

Share this post


Link to post
Share on other sites

https://www.nytimes.com/2018/01/02/magazine/can-an-algorithm-tell-when-kids-are-in-danger.html?mc=aud_dev&mcid=fb-nytimes&mccr=JanMC34&mcdt=2018-01&subid=JanMC34&ad-keywords=AudDevGate

 

I saw a New York Times article recently about that from a different perspective. I will have to read the one you linked later.

It's a good article, presents pros and cons well. Thx!

Share this post


Link to post
Share on other sites

I have been reading the threads on child poverty with deep interest and sorrow at what some members here have gone through.  Today I happened to read an article in Wired magazine called "A Child Abuse Prediction Model Fails Poor Families".  

 

Basically, Allegheny County, PA is using this model to predict child abuse, but the model uses data mined from public service records, such as child welfare, Head Start, county housing authority, drug and alcohol services-pretty much all services that only poor people use--in essence, reporting "how many public resources families have consumed", and using that to predict child abuse risk.

 

Here was the paragraph that really struck a chord in me:  

 

"We might call this poverty profiling. Like racial profiling, poverty profiling targets individuals for extra scrutiny based not on their behavior but rather on a personal characteristic: They live in poverty. Because the model confuses parenting while poor with poor parenting, the AFST views parents who reach out to public programs as risks to their children."

 

Anyway, this just seems like one more thing designed to help people that in actuality bites them in the rear.  The author of the article suggested averaging the model's score with one a human worker comes up with, but that doesn't happen, and it doesn't seem like it ever will.  Machines are always right, and certainly not prejudiced, of course!

 

Garbage in, garbage out.

 

Its sounds like it was a start of a model, but a lot more work needs to be done.  Me, I would pick sociopath parent and insistence on isolating the children as two of the characteristics.  Might also add isolating the mother.

Share this post


Link to post
Share on other sites

It's yet another no-win situation.

 

We can say that not every child in poverty is abused and not every abused child is in poverty and be correct, but there is a known relationship, even though it isn't 1:1. If we target these families for closer scrutiny, we mess with innocent families and probably create an extra barrier to help for *all families. But, if we ignore the relationship that does exist, aren't we knowingly making it harder to reduce abuse?

 

The conversations going on right now weigh heavily on me... as a homeschooler, as someone who's been investigated (for non-homeschool-related false accusations), as someone with young relatives who are being failed by adults, in both the family and the system, as someone with other relatives who work in the system, as someone who lives in a high poverty area with many serious issues, and as someone who's lived at or near poverty level in the past with zero abuse.

 

I'm not doing anything wrong and I don't want to be messed with. That's true. But SOMETHING needs to be done.

 

Like the whole school reform thing, if we sit around waiting for the perfect solution, we're going to lose kids.

How about we put our efforts into reducing poverty, thus reducing stress on families and abuse of children.

 

Win win win except for those who want to amass and horde all the wealth for themselves.

  • Like 3

Share this post


Link to post
Share on other sites

How about we put our efforts into reducing poverty, thus reducing stress on families and abuse of children.

 

Win win win except for those who want to amass and horde all the wealth for themselves.

 

I'm 100% behind that.  But, again,

 

 

 

Like the whole school reform thing, if we sit around waiting for the perfect solution, we're going to lose kids.

 

If we can get every single person to work toward eradicating poverty and the lingering effects of some poverty, it will still take a lot of time.

Share this post


Link to post
Share on other sites

It reminds me of when older dd was born. Dh had lost his job and I had state medical assistance. The county must do some kind of poverty-based targeting, because I had every agency calling me to see if I needed anything (which was nice) and public health nurses constantly calling and stopping by and mailing me things about how not to be a crappy parent (which was not so nice). The assumption that I was going to abuse or neglect my child, and that public health needed to protect my new baby from me and my parenting, really hurt my self esteem and my self image as a new parent, especially when I ended up with PPA. It took me a while to convince myself that I wasn't an awful parent just because we were poor at the time. It really did shape how I saw myself in those early days, and not in a good way.

  • Like 8

Share this post


Link to post
Share on other sites

Yikes. This is why correlation science can often times run amok. Looking at large swaths of statistical data one can draw correlations with many things.

 

The real issue with why poverty precipitates out in this way, unfortunately, is because you will find a higher number of people who are poor and who abuse. Often this isn't due to the poverty of course, it is due to poor self control (issues with executive functioning, impulsivity, reactivity etc) which might lead to both being poor and being abusive. Much like the propensity to abuse, being poor is a factor of that struggle with self control and not due to being poor in and of itself. So basically, more people who abuse will be poor because being reactive and impulsive or having low grit and perserverence does not often lead to high educational attainment or the steps necessary to love into a higher income bracket. However, this does not work the other way around (people who are poor are not more likely to abuse) because being poor isn't always due to an inability to delay gratification long enough to get ahead. Sometimes it is a choice to choose a low paying career, sometimes it is health related, sometimes it is sacrifice to be with the children...this is not a situation that often leads to abuse. Trying to take human behavior and distilling it down into a mathematical function is just not helpful at all. This is exactly how biases and prejudices begin to precipitate within a culture. It is gross and strips people of their dignity.

 

While I would agree about your arguments about impulse control and abuse, I believe you are conflating a lot of factors with regard to poverty. Since that's where you start, it gives the appearance that that is what you believe the main cause of poverty to be. Is that accurate? 

 

WIth regard to CPS referrals, this line of argument omits  the fact that CPS referrals include charges of neglect, not just abuse. 

Neglect includes things like not having enough food in the refrigerator when a social worker comes to inspect (how might that be related to poverty?, not having heat in the home (how might that be related to poverty? leaving the kids home alone while the parent works ( Your argument appears to assume that poverty is a result of not working, yet a parent could work full time at a minimum wage job and be below the poverty line. How will they also pay for babysitting?)  Many families are one  job loss away from  poverty level living. Furthermore, just as wealth tends to beget wealth, poverty tends to beget poverty. Had our family  ever been in bad straits, due to loss of income, etc. we have two sets of grandparents who could have supported us. Many people who live in poverty have relatives in the same boat, so there is no family safety net either.  Or there are one or two family members who were able to climb out and they keep their distance because they can't help everyone and need to be their own safety net. 

 

 In the article, the second description (of the 14 year old) sounds like pure poverty and someone being reported because they are poor. If the social worker finds that poverty is the reason for the conditions (and not for instance, that the parents eat fine but their children don't) then she's supposed to refer them to ta-dah: the agencies in the community that can help with income issues such as ones that provide heat in the winter, food stamps, etc. But if the family follows through on that, their "risk score" increases!  

 

And the points that middle class families may have the same needs but be getting them met privately (through private mental health, financial supplementation from families, etc.) means this is absolutely a discriminatory tool that will raise the risk profile and the likelihood of a CPS investigation of those who are poor.  The article is one of the few that note that a CPS investigation is not benign and can have devastating and permanent effects even if in the end no abuse is substantiated. 

Edited by Laurie4b
  • Like 3

Share this post


Link to post
Share on other sites

It reminds me of when older dd was born. Dh had lost his job and I had state medical assistance. The county must do some kind of poverty-based targeting, because I had every agency calling me to see if I needed anything (which was nice) and public health nurses constantly calling and stopping by and mailing me things about how not to be a crappy parent (which was not so nice). The assumption that I was going to abuse or neglect my child, and that public health needed to protect my new baby from me and my parenting, really hurt my self esteem and my self image as a new parent, especially when I ended up with PPA. It took me a while to convince myself that I wasn't an awful parent just because we were poor at the time. It really did shape how I saw myself in those early days, and not in a good way.

 

This is where a service for everyone can help.  In the UK, midwives and then health visitors visit mother and child for a couple of weeks after the birth.  You can refuse, but most people don't.  Because almost everyone receives the visits, there's no stigma, and it can be helpful.

  • Like 6

Share this post


Link to post
Share on other sites

Probably, but on the other hand, we're reinforcing the class privilege of middle and upper class abusers. They're respectable, so they're given a pass on things poorer people don't get a pass on. btdt

 

Yep.

 

Abuse happens in all socio-economic classes.

Share this post


Link to post
Share on other sites

Yep.

 

Abuse happens in all socio-economic classes.

I've wondered how our assumptions about class and abuse impacted the lack of concern in the 13 children case. Dad's a well educated engineer with $140,000/year income.

  • Like 2

Share this post


Link to post
Share on other sites

Honestly the worst abuse cases we've had or known about in terms of severity were from families with two straight married parents where at least one parent had a good job.  They were coherent and sophisticated enough to say all the right things to investigators, and yet they kept getting reported for abuse. I think it's more difficult to investigate cases like that. 

  • Like 1

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

SUBSCRIBE TO OUR NEWSLETTER & RECEIVE A COUPON FOR
10% OFF
We respect your privacy.You’ll hear about new products, special discounts & sales, and homeschooling tips. *Coupon only valid for first-time registrants. Coupon cannot be combined with any other offer. Entering your email address makes you eligible to receive future promotional emails.
0 Shares
Share
Tweet
Pin
×