Sergio, the idea you describe in your first paragraph has been formalized
with this algorithm <http://projects.iq.harvard.edu/frontier/home>, and so
that's another option. With CEM, you would decide how important it is to
get matches for each variable and coursen more for less important
variables.
Gary
---
http://gking.harvard.edu
617-500-7570
On Wed, Apr 5, 2017 at 12:32 PM, Sergio Salis <Sergio.Salis(a)natcen.ac.uk>
wrote:
> Hi Gary,
>
>
>
> Thanks very much for your advice. I understand the idea is trying
> different coarsening strategies (among those which make sense) for each
> variable and see which one produces the lowest imbalance, measured by means
> of the Multivariate L1 distance (the univariate imbalances should also be
> looked at individually). Is this correct?
>
>
>
> For variables like income and assets I guess it makes sense to use
> percentiles as there is no obvious value to create cut-off points. If so,
> shall I use
>
>
>
> cem var1 var 2 …. income(P1 P2 …. Pn) , treatment(treated)
>
>
>
> (where P1=value of the 1st percentile, P2=value of the 2nd percentile …..
> Pn=value of the last percentile)?
>
>
>
> (#10) will produce 10 equally sized bins but I am not sure whether equal
> size means equal base (e.g. bin 1 includes those with income between 1 to
> 10, bin 2 those with income 11 to 20, etc.) or equal frequencies (in which
> case a bin defines a percentile). I am also not sure what Sturge's rule and
> Scott’s algorithm are, I cannot find any description in the Stata help file.
>
>
>
> Thanks again for your help< very much appreciated.
>
>
>
> Sergio
>
>
>
> *From:* Gary King [mailto:king@harvard.edu]
> *Sent:* 05 April 2017 16:06
> *To:* Sergio Salis
> *Cc:* cem(a)lists.gking.harvard.edu
> *Subject:* Re: [cem] Few matched strata (and individuals within strata)
> after implementing cem
>
>
>
> Hi Sergio, you can adjust the coarsening rather than using the defaults in
> CEM. more coarse bins will generate more observations. you want to make
> the choices based on the substance of the variables, and which ones are
> more important to match finely on
>
>
> Gary
>
> --
>
> *Gary King* - Albert J. Weatherhead III University Professor - Director,
> IQSS <http://iq.harvard.edu/> - Harvard University
>
> GaryKing.org - King(a)Harvard.edu - @KingGary <https://twitter.com/kinggary> -
> 617-500-7570 <(617)%20500-7570> - Assistant <king-assist(a)iq.harvard.edu>:
> 617-495-9271 <(617)%20495-9271>
>
>
>
> On Wed, Apr 5, 2017 at 10:04 AM, Sergio Salis <Sergio.Salis(a)natcen.ac.uk>
> wrote:
>
>
>
> Hi all,
>
>
>
> I’m considering using the cem Stata programme to evaluate the impact of a
> welfare-to-work programme in the UK. However, I have never used cem before
> so I am trying to understand some basic issues before proceeding with the
> estimation.
>
>
>
> The first thing I’d be interested in understanding is: How does one handle
> situations where after running cem the number of matched strata (and units
> within them) are very small?
>
>
>
> Applying the cem algorithm to data from a previous impact evaluation I get:
>
>
>
> Number of strata: 8883
>
> Number of matched strata: 132
>
>
>
> 0 1
>
> All 8208 1584
>
> Matched 179 141
>
> Unmatched 8029 1443
>
>
>
> If I calculate the ATT using cem matched data I get an impact estimate
> which is positive (around 5ppts; based on 320 obs only) while using
> psmatch2 on all data (i.e. not only those in cem matched strata; around
> 8,237 obs are used) with kernel weights I get an estimate of around
> -5.7ppts. This means I reach opposite conclusions about the impact of the
> programme of interest using cem and psmatch2.
>
>
>
> I understand the cem-based estimates are based on better matched data
> (i.e. produce less biased estimates) compared to my psmatch2 estimate with
> kernel weights) but this comes at the expense of external validity:
> inference on the initial population is made based on a very small subset of
> data (estimates based on cem are not statistically significant while my
> original estimate was highly significant). Any advice about how one can
> handle situations of this type?
>
>
>
> Many thanks,
>
> Sergio
>
>
>
>
>
>
>
> NatCen Social Research
> 35 Northampton Square
> London EC1V 0AX
> 020 7250 1866
>
> Visit our website. www.natcen.ac.uk
> Read our latest blog. http://www.natcen.ac.uk/blog
> Follow us. @NatCen <https://twitter.com/natcen>
> Email us. info(a)natcen.ac.uk
>
> NatCen Social Research is certificated to ISO/IEC 27001:2013 for
> Information Security Management Systems and to ISO 20252:2012, the
> international standard for market, opinion and social research.
>
> Company limited by guarantee. Registered in England No. 4392418. Charity
> registered in England and Wales (1091768) and in Scotland (SC038454).
>
> Confidentiality: The information in this email and any attachments are
> confidential and may include some that is legally privileged. It must not
> be disclosed to or used by persons other than the intended recipient. If
> received in error, please notify us immediately and then delete this
> document.
> Content: Any views or opinions expressed do not necessarily represent
> those of NatCen Social Research. Please note the content of this e-mail may
> be intercepted, monitored or recorded for compliance purposes. Sensitive
> personal data should not normally be transmitted by e-mail.
> Copyright: Copyright in this e-mail and any attachments created by NatCen
> Social Research belong to NatCen Social Research unless otherwise stated.
> Care: NatCen Social Research shall not be liable to the recipient or any
> third party for any loss or damage howsoever arising from this e-mail
> and/or its content, including loss or damage caused by virus. It is the
> responsibility of the recipient to ensure the opening or use of this
> message and any attachments shall not adversely affect systems or data.
>
>
>
Sergio, the idea you describe in your first paragraph has been formalized
with this algorithm <http://projects.iq.harvard.edu/frontier/home>, and so
that's another option. With CEM, you would decide how important it is to
get matches for each variable and coursen more for less important
variables.
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS <http://iq.harvard.edu/> - Harvard University
GaryKing.org - King(a)Harvard.edu - @KingGary <https://twitter.com/kinggary> -
617-500-7570 - Assistant <king-assist(a)iq.harvard.edu>: 617-495-9271
On Wed, Apr 5, 2017 at 12:32 PM, Sergio Salis <Sergio.Salis(a)natcen.ac.uk>
wrote:
> Hi Gary,
>
>
>
> Thanks very much for your advice. I understand the idea is trying
> different coarsening strategies (among those which make sense) for each
> variable and see which one produces the lowest imbalance, measured by means
> of the Multivariate L1 distance (the univariate imbalances should also be
> looked at individually). Is this correct?
>
>
>
> For variables like income and assets I guess it makes sense to use
> percentiles as there is no obvious value to create cut-off points. If so,
> shall I use
>
>
>
> cem var1 var 2 …. income(P1 P2 …. Pn) , treatment(treated)
>
>
>
> (where P1=value of the 1st percentile, P2=value of the 2nd percentile …..
> Pn=value of the last percentile)?
>
>
>
> (#10) will produce 10 equally sized bins but I am not sure whether equal
> size means equal base (e.g. bin 1 includes those with income between 1 to
> 10, bin 2 those with income 11 to 20, etc.) or equal frequencies (in which
> case a bin defines a percentile). I am also not sure what Sturge's rule and
> Scott’s algorithm are, I cannot find any description in the Stata help file.
>
>
>
> Thanks again for your help< very much appreciated.
>
>
>
> Sergio
>
>
>
> *From:* Gary King [mailto:king@harvard.edu]
> *Sent:* 05 April 2017 16:06
> *To:* Sergio Salis
> *Cc:* cem(a)lists.gking.harvard.edu
> *Subject:* Re: [cem] Few matched strata (and individuals within strata)
> after implementing cem
>
>
>
> Hi Sergio, you can adjust the coarsening rather than using the defaults in
> CEM. more coarse bins will generate more observations. you want to make
> the choices based on the substance of the variables, and which ones are
> more important to match finely on
>
>
> Gary
>
> --
>
> *Gary King* - Albert J. Weatherhead III University Professor - Director,
> IQSS <http://iq.harvard.edu/> - Harvard University
>
> GaryKing.org - King(a)Harvard.edu - @KingGary <https://twitter.com/kinggary> -
> 617-500-7570 <(617)%20500-7570> - Assistant <king-assist(a)iq.harvard.edu>:
> 617-495-9271 <(617)%20495-9271>
>
>
>
> On Wed, Apr 5, 2017 at 10:04 AM, Sergio Salis <Sergio.Salis(a)natcen.ac.uk>
> wrote:
>
>
>
> Hi all,
>
>
>
> I’m considering using the cem Stata programme to evaluate the impact of a
> welfare-to-work programme in the UK. However, I have never used cem before
> so I am trying to understand some basic issues before proceeding with the
> estimation.
>
>
>
> The first thing I’d be interested in understanding is: How does one handle
> situations where after running cem the number of matched strata (and units
> within them) are very small?
>
>
>
> Applying the cem algorithm to data from a previous impact evaluation I get:
>
>
>
> Number of strata: 8883
>
> Number of matched strata: 132
>
>
>
> 0 1
>
> All 8208 1584
>
> Matched 179 141
>
> Unmatched 8029 1443
>
>
>
> If I calculate the ATT using cem matched data I get an impact estimate
> which is positive (around 5ppts; based on 320 obs only) while using
> psmatch2 on all data (i.e. not only those in cem matched strata; around
> 8,237 obs are used) with kernel weights I get an estimate of around
> -5.7ppts. This means I reach opposite conclusions about the impact of the
> programme of interest using cem and psmatch2.
>
>
>
> I understand the cem-based estimates are based on better matched data
> (i.e. produce less biased estimates) compared to my psmatch2 estimate with
> kernel weights) but this comes at the expense of external validity:
> inference on the initial population is made based on a very small subset of
> data (estimates based on cem are not statistically significant while my
> original estimate was highly significant). Any advice about how one can
> handle situations of this type?
>
>
>
> Many thanks,
>
> Sergio
>
>
>
>
>
>
>
> NatCen Social Research
> 35 Northampton Square
> London EC1V 0AX
> 020 7250 1866
>
> Visit our website. www.natcen.ac.uk
> Read our latest blog. http://www.natcen.ac.uk/blog
> Follow us. @NatCen <https://twitter.com/natcen>
> Email us. info(a)natcen.ac.uk
>
> NatCen Social Research is certificated to ISO/IEC 27001:2013 for
> Information Security Management Systems and to ISO 20252:2012, the
> international standard for market, opinion and social research.
>
> Company limited by guarantee. Registered in England No. 4392418. Charity
> registered in England and Wales (1091768) and in Scotland (SC038454).
>
> Confidentiality: The information in this email and any attachments are
> confidential and may include some that is legally privileged. It must not
> be disclosed to or used by persons other than the intended recipient. If
> received in error, please notify us immediately and then delete this
> document.
> Content: Any views or opinions expressed do not necessarily represent
> those of NatCen Social Research. Please note the content of this e-mail may
> be intercepted, monitored or recorded for compliance purposes. Sensitive
> personal data should not normally be transmitted by e-mail.
> Copyright: Copyright in this e-mail and any attachments created by NatCen
> Social Research belong to NatCen Social Research unless otherwise stated.
> Care: NatCen Social Research shall not be liable to the recipient or any
> third party for any loss or damage howsoever arising from this e-mail
> and/or its content, including loss or damage caused by virus. It is the
> responsibility of the recipient to ensure the opening or use of this
> message and any attachments shall not adversely affect systems or data.
>
>
>
Hi Sergio, you can adjust the coarsening rather than using the defaults in
CEM. more coarse bins will generate more observations. you want to make
the choices based on the substance of the variables, and which ones are
more important to match finely on
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS <http://iq.harvard.edu/> - Harvard University
GaryKing.org - King(a)Harvard.edu - @KingGary <https://twitter.com/kinggary> -
617-500-7570 - Assistant <king-assist(a)iq.harvard.edu>: 617-495-9271
On Wed, Apr 5, 2017 at 10:04 AM, Sergio Salis <Sergio.Salis(a)natcen.ac.uk>
wrote:
>
> Hi all,
>
>
>
> I’m considering using the cem Stata programme to evaluate the impact of a
> welfare-to-work programme in the UK. However, I have never used cem before
> so I am trying to understand some basic issues before proceeding with the
> estimation.
>
>
>
> The first thing I’d be interested in understanding is: How does one handle
> situations where after running cem the number of matched strata (and units
> within them) are very small?
>
>
>
> Applying the cem algorithm to data from a previous impact evaluation I get:
>
>
>
> Number of strata: 8883
>
> Number of matched strata: 132
>
>
>
> 0 1
>
> All 8208 1584
>
> Matched 179 141
>
> Unmatched 8029 1443
>
>
>
> If I calculate the ATT using cem matched data I get an impact estimate
> which is positive (around 5ppts; based on 320 obs only) while using
> psmatch2 on all data (i.e. not only those in cem matched strata; around
> 8,237 obs are used) with kernel weights I get an estimate of around
> -5.7ppts. This means I reach opposite conclusions about the impact of the
> programme of interest using cem and psmatch2.
>
>
>
> I understand the cem-based estimates are based on better matched data
> (i.e. produce less biased estimates) compared to my psmatch2 estimate with
> kernel weights) but this comes at the expense of external validity:
> inference on the initial population is made based on a very small subset of
> data (estimates based on cem are not statistically significant while my
> original estimate was highly significant). Any advice about how one can
> handle situations of this type?
>
>
>
> Many thanks,
>
> Sergio
>
>
>
>
>
> NatCen Social Research
> 35 Northampton Square
> London EC1V 0AX
> 020 7250 1866
>
> Visit our website. www.natcen.ac.uk
> Read our latest blog. http://www.natcen.ac.uk/blog
> Follow us. @NatCen <https://twitter.com/natcen>
> Email us. <info(a)natcen.ac.uk>info(a)natcen.ac.uk
>
> NatCen Social Research is certificated to ISO/IEC 27001:2013 for
> Information Security Management Systems and to ISO 20252:2012, the
> international standard for market, opinion and social research.
>
> Company limited by guarantee. Registered in England No. 4392418. Charity
> registered in England and Wales (1091768) and in Scotland (SC038454).
>
> Confidentiality: The information in this email and any attachments are
> confidential and may include some that is legally privileged. It must not
> be disclosed to or used by persons other than the intended recipient. If
> received in error, please notify us immediately and then delete this
> document.
> Content: Any views or opinions expressed do not necessarily represent
> those of NatCen Social Research. Please note the content of this e-mail may
> be intercepted, monitored or recorded for compliance purposes. Sensitive
> personal data should not normally be transmitted by e-mail.
> Copyright: Copyright in this e-mail and any attachments created by NatCen
> Social Research belong to NatCen Social Research unless otherwise stated.
> Care: NatCen Social Research shall not be liable to the recipient or any
> third party for any loss or damage howsoever arising from this e-mail
> and/or its content, including loss or damage caused by virus. It is the
> responsibility of the recipient to ensure the opening or use of this
> message and any attachments shall not adversely affect systems or data.
>
Hi all,
I'm considering using the cem Stata programme to evaluate the impact of a welfare-to-work programme in the UK. However, I have never used cem before so I am trying to understand some basic issues before proceeding with the estimation.
The first thing I'd be interested in understanding is: How does one handle situations where after running cem the number of matched strata (and units within them) are very small?
Applying the cem algorithm to data from a previous impact evaluation I get:
Number of strata: 8883
Number of matched strata: 132
0 1
All 8208 1584
Matched 179 141
Unmatched 8029 1443
If I calculate the ATT using cem matched data I get an impact estimate which is positive (around 5ppts; based on 320 obs only) while using psmatch2 on all data (i.e. not only those in cem matched strata; around 8,237 obs are used) with kernel weights I get an estimate of around -5.7ppts. This means I reach opposite conclusions about the impact of the programme of interest using cem and psmatch2.
I understand the cem-based estimates are based on better matched data (i.e. produce less biased estimates) compared to my psmatch2 estimate with kernel weights) but this comes at the expense of external validity: inference on the initial population is made based on a very small subset of data (estimates based on cem are not statistically significant while my original estimate was highly significant). Any advice about how one can handle situations of this type?
Many thanks,
Sergio
NatCen Social Research
35 Northampton Square
London EC1V 0AX
020 7250 1866
Visit our website. www.natcen.ac.uk
Read our latest blog. natcenblog.blogspot.com
Follow us. @NatCen
Email us. info(a)natcen.ac.uk
NatCen Social Research is certificated to ISO/IEC 27001:2013 for Information Security Management Systems and to ISO 20252:2012, the international standard for market, opinion and social research.
Company limited by guarantee. Registered in England No. 4392418. Charity registered in England and Wales (1091768) and in Scotland (SC038454).
Confidentiality: The information in this email and any attachments are confidential and may include some that is legally privileged. It must not be disclosed to or used by persons other than the intended recipient. If received in error, please notify us immediately and then delete this document.
Content: Any views or opinions expressed do not necessarily represent those of NatCen Social Research. Please note the content of this e-mail may be intercepted, monitored or recorded for compliance purposes. Sensitive personal data should not normally be transmitted by e-mail.
Copyright: Copyright in this e-mail and any attachments created by NatCen Social Research belong to NatCen Social Research unless otherwise stated.
Care: NatCen Social Research shall not be liable to the recipient or any third party for any loss or damage howsoever arising from this e-mail and/or its content, including loss or damage caused by virus. It is the responsibility of the recipient to ensure the opening or use of this message and any attachments shall not adversely affect systems or data.
Hi everyone,
I have two questions:
Question 1. I created this sample dataset (test):
code Age open outcome
1 A 12 0 1
2 B 15 0 0
3 C 18 0 1
4 D 12 1 0
5 E 18 1 1
6 F 20 1 0
When I run this command:
todrop <- c("outcome", "code")
cem2 <- cem (treatment = "open", data = test, drop = todrop , k2k=TRUE)
I get this data back :
code Age open outcome
<chr> <dbl> <dbl> <dbl>
1 A 12 0 1
2 C 18 0 1
3 D 12 1 0
4 F 20 1 0
When I use matchit
match <- matchit(open ~ Age, test, method = "exact")
I get this result
code Age open outcome weights subclass
1 A 12 0 1 1 1
3 C 18 0 1 1 2
4 D 12 1 0 1 1
5 E 18 1 1 1 2
So, my question is why CEM does not chose the record "E" with age 18 and
chooses the one with age 20. Is the exact method in matchit more accurate
than CEM in this case?
Question 2. I have a database with 140k records and 440 variables, which I
want to match on only 20 variables. If I want to use CEM, is there an easy
way to include those 20 variables, and not drop the other 420?
Thanks a lot in advance.
-Ashkan
Hi,
I am trying to install CEM for SPSS 24. The most updated version is for
V23. I receive an error message "SPSS not found. Aborting installation". Am
I doing something wrong or the versions won't match?
Best,
-Ashkan
Hello,
I am looking into using CEM in a context where I need to match subsets of
the treatment group separately based on the timing of treatment
(unfortunately, the simple solution of using time as a matching component
won't work in my case).
After matching, I would like to combine the matched subsets to estimate a
pooled treatment effect. Do the cem_weights need to be revised to account
for pooling? If so, is there a reference that describes how the cem_weights
are created? I understand their basic purpose - to account for differential
strata sizes - but I'm not clear on the actual formula that is used to
generate the weights for matched controls.
Thank you,
Zack
Hi Catherine, thanks for your note to the list. It sounds like you could
define this as at one level but with multiple (rather than binary)
treatment regimes. Our papers on cem explain how that works. Best of luck
with your research. Gary King
---
GaryKing.org
617-500-7570
On Dec 2, 2016 4:49 PM, "Catherine E Hendrick" <emily.hendrick(a)utexas.edu>
wrote:
I am considering using Coarsened Exact Matching in order to study the
effects of teen mothers' degree type (GED v. HS diploma) on long-term
outcomes. In order to sufficiently address my research questions, it seems
that I may need two levels of matching:
1) matching those who attained a GED with those who attained a HS diploma
(to isolate the effects of degree type on outcomes)
2) matching teen mothers with women who began childbearing after the
teenage years (to determine if diploma type effects on later outcomes are
the same or different for women who began childbearing during the teenage
years v. later)
I haven't seen any literature using CEM that conducts two levels of
matching as I am proposing. Do you know of literature where researchers
have done this and/or do you see any methodological reason for NOT
conducting two levels of matching, as I have proposed, when using CEM?
Many thanks for any information you're able to pass along.
Dear Catherine E Hendrick,
I think your problem can be easily re-conducted to the case of Multiple treatment doses (e.g., doses of a drug).
Among other see section
6.1.4 of Stuart (2010) paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2943670/
Stat Sci. 2010 Feb 1; 25(1): 1–21. doi: 10.1214/09-STS313
I was using CEM to pairwise compare outcomes of three groups of firms with different treatments (degree of environmental commitment ) http://www.sciencedirect.com/science/article/pii/S0959652616313804 but I miss a formal/ theoretical discussion and I will be very interested on proper guidance on this point.
Cesare
From: "cem-request(a)lists.gking.harvard.edu" <cem-request(a)lists.gking.harvard.edu>
To: cem(a)lists.gking.harvard.edu
Sent: Saturday, 3 December 2016, 18:00
Subject: Cem Digest, Vol 83, Issue 1
Send Cem mailing list submissions to
cem(a)lists.gking.harvard.edu
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.gking.harvard.edu/mailman/listinfo/cem
or, via email, send a message with subject or body 'help' to
cem-request(a)lists.gking.harvard.edu
You can reach the person managing the list at
cem-owner(a)lists.gking.harvard.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Cem digest..."
Today's Topics:
1. 2-level CEM? (Catherine E Hendrick)
----------------------------------------------------------------------
Message: 1
Date: Fri, 2 Dec 2016 15:49:08 -0600
From: Catherine E Hendrick <emily.hendrick(a)utexas.edu>
To: cem(a)lists.gking.harvard.edu
Subject: [cem] 2-level CEM?
Message-ID:
<CAJWC3AAG_ejC3Y2S1FUo6aG2DJ4RFnSCpNh=npijS-PoNCqmbA(a)mail.gmail.com>
Content-Type: text/plain; charset="utf-8"
I am considering using Coarsened Exact Matching in order to study the
effects of teen mothers' degree type (GED v. HS diploma) on long-term
outcomes. In order to sufficiently address my research questions, it seems
that I may need two levels of matching:
1) matching those who attained a GED with those who attained a HS diploma
(to isolate the effects of degree type on outcomes)
2) matching teen mothers with women who began childbearing after the
teenage years (to determine if diploma type effects on later outcomes are
the same or different for women who began childbearing during the teenage
years v. later)
I haven't seen any literature using CEM that conducts two levels of
matching as I am proposing. Do you know of literature where researchers
have done this and/or do you see any methodological reason for NOT
conducting two levels of matching, as I have proposed, when using CEM?
Many thanks for any information you're able to pass along.
I am considering using Coarsened Exact Matching in order to study the
effects of teen mothers' degree type (GED v. HS diploma) on long-term
outcomes. In order to sufficiently address my research questions, it seems
that I may need two levels of matching:
1) matching those who attained a GED with those who attained a HS diploma
(to isolate the effects of degree type on outcomes)
2) matching teen mothers with women who began childbearing after the
teenage years (to determine if diploma type effects on later outcomes are
the same or different for women who began childbearing during the teenage
years v. later)
I haven't seen any literature using CEM that conducts two levels of
matching as I am proposing. Do you know of literature where researchers
have done this and/or do you see any methodological reason for NOT
conducting two levels of matching, as I have proposed, when using CEM?
Many thanks for any information you're able to pass along.