Hi
Bruce. I got both copies and enjoyed reading this both times.
Alice
I sent
this post this morning but have no indication that it actually appeared
and therefore am resending it. My apology if it’s a
duplicate.
Bruce
···
From:
Bruce Abbott [mailto:bbabbott@frontier.com] Sent:
Saturday, October 11, 2014 9:12 AM To: CSGnet Subject: Casting Nets and Testing Specimens
[From Bruce Abbott
(2014.10.11.0910 EDT)]
November’s Discover magazine contains an article entitled “Singled
Out” that points out a problem with the “gold standard”
in medical research, the randomized control clinical trial used to
evaluate the efficacy of drugs and other therapeutic measures. According
to the article, “ . . . even the gold standard isn’t perfect. The controlled clinical trial is really about averages, and averages
don’t necessarily tell you what will happen to an individual. Such
a trial might tell you that, statistically speaking, milk isn’t good
for Crohn’s patients. But within that sample, there might be people
who didn’t have any problems drinking milk, and people whose
symptoms even got better while drinking it.”
This problem
isn’t news for those of us in PCT – it is one that Bill Powers
railed against for decades and is the thesis of Phil Runkel’s (1990)
book, Casting Nets and Testing Specimens. Phil argued that
large-N, group-based studies, with their random assignment of participants
to treatments, statistical averaging, and inferential statistical analyses
to estimate the reliability of any average differences between groups, are
useful as screening tools for identifying potentially important
relationships (“casting nets”), but when it comes down to
treating individuals, one must test individuals individually
(“testing specimens”), for the very reason noted in the quoted
paragraph. PCT methodology, of course, does exactly that, as when
performing “the test for the controlled variable.”
The Discover
article traces the history of attempts to apply this
“single-subject” or “N of 1” approach when
evaluating the effects of medical treatments on individual patients. “One way to correct for the gaps the gold standard leaves in our
knowledge is the “N of 1” trial, where the number of
participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis –
for example, “I think drinking milk will make me feel sick. Am I
right?” After a rocky start, the N of 1 approach is finally coming
into its own under the banner of personalized medicine. “Single-subject trials allow you to personalize the outcome, not
just the treatment. . . . That ability to improve the quality of care in a
truly personalized way explains why N of 1 trials are making a
comeback.”
It’s an approach
that’s been at the heart of PCT from the beginning. It’s nice
to see that the medical world may finally be catching
up.
November’s Discover magazine contains an article entitled “Singled Out” that points out a problem with the “gold standard” in medical research, the randomized control clinical trial used to evaluate the efficacy of drugs and other therapeutic measures. According to the article, “ . . . even the gold standard isn’t perfect. The controlled clinical trial is really about averages, and averages don’t necessarily tell you what will happen to an individual. Such a trial might tell you that, statistically speaking, milk isn’t good for Crohn’s patients. But within that sample, there might be people who didn’t have any problems drinking milk, and people whose symptoms even got better while drinking it.”
This problem isn’t news for those of us in PCT – it is one that Bill Powers railed against for decades and is the thesis of Phil Runkel’s (1990) book, Casting Nets and Testing Specimens. Phil argued that large-N, group-based studies, with their random assignment of participants to treatments, statistical averaging, and inferential statistical analyses to estimate the reliability of any average differences between groups, are useful as screening tools for identifying potentially important relationships (“casting nets”), but when it comes down to treating individuals, one must test individuals individually (“testing specimens”), for the very reason noted in the quoted paragraph. PCT methodology, of course, does exactly that, as when performing “the test for the controlled variable.”
The Discover article traces the history of attempts to apply this “single-subject” or “N of 1” approach when evaluating the effects of medical treatments on individual patients. “One way to correct for the gaps the gold standard leaves in our knowledge is the “N of 1” trial, where the number of participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis – for example, “I think drinking milk will make me feel sick. Am I right?” After a rocky start, the N of 1 approach is finally coming into its own under the banner of personalized medicine. “Single-subject trials allow you to personalize the outcome, not just the treatment. . . . That ability to improve the quality of care in a truly personalized way explains why N of 1 trials are making a comeback.”
It’s an approach that’s been at the heart of PCT from the beginning. It’s nice to see that the medical world may finally be catching up.
I sent this post this morning but have no indication that it actually appeared and therefore am resending it. My apology if it’s a duplicate.
Bruce
···
From: Bruce Abbott [mailto:bbabbott@frontier.com] Sent: Saturday, October 11, 2014 9:12 AM To: CSGnet Subject: Casting Nets and Testing Specimens
[From Bruce Abbott (2014.10.11.0910 EDT)]
November’s Discover magazine contains an article entitled “Singled Out” that points out a problem with the “gold standard” in medical research, the randomized control clinical trial used to evaluate the efficacy of drugs and other therapeutic measures. According to the article, “ . . . even the gold standard isn’t perfect. The controlled clinical trial is really about averages, and averages don’t necessarily tell you what will happen to an individual. Such a trial might tell you that, statistically speaking, milk isn’t good for Crohn’s patients. But within that sample, there might be people who didn’t have any problems drinking milk, and people whose symptoms even got better while drinking it.”
This problem isn’t news for those of us in PCT – it is one that Bill Powers railed against for decades and is the thesis of Phil Runkel’s (1990) book, Casting Nets and Testing Specimens. Phil argued that large-N, group-based studies, with their random assignment of participants to treatments, statistical averaging, and inferential statistical analyses to estimate the reliability of any average differences between groups, are useful as screening tools for identifying potentially important relationships (“casting nets”), but when it comes down to treating individuals, one must test individuals individually (“testing specimens”), for the very reason noted in the quoted paragraph. PCT methodology, of course, does exactly that, as when performing “the test for the controlled variable.”
The Discover article traces the history of attempts to apply this “single-subject” or “N of 1” approach when evaluating the effects of medical treatments on individual patients. “One way to correct for the gaps the gold standard leaves in our knowledge is the “N of 1” trial, where the number of participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis – for example, “I think drinking milk will make me feel sick. Am I right?” After a rocky start, the N of 1 approach is finally coming into its own under the banner of personalized medicine. “Single-subject trials allow you to personalize the outcome, not just the treatment. . . . That ability to improve the quality of care in a truly personalized way explains why N of 1 trials are making a comeback.”
It’s an approach that’s been at the heart of PCT from the beginning. It’s nice to see that the medical world may finally be catching up.
BA: Novemberâs Discover magazine contains an article entitled âSingled Outâ? that points out a problem with the âgold standardâ? in medical research, the randomized control clinical trial used to evaluate the efficacy of drugs and other therapeutic measures. According to the article, â . . . even the gold standard isnât perfect. The controlled clinical trial is really about averages, and averages donât necessarily tell you what will happen to an individual. Such a trial might tell you that, statistically speaking, milk isnât good for Crohnâs patients. But within that sample, there might be people who didnât have any problems drinking milk, and people whose symptoms even got better while drinking it.â?
RM: I think it’s important to be clear that the problem with this kind of research – what Runkel called “Casting Nets” research – is the use of group data to come to conclusions about individuals. The PCT objection to “Casting Nets” research is that it is the main way research is done in psychology (with notable exceptions in the operant conditioning and perception areas) where the goal is presumably to understand the nature of the individual; actually, what are the psychological “laws” that apply to all individuals, taken one at a time.Â
RM: I think the “Casting Nets” approach to research is appropriate when the object of study is a group rather than an individual. In particular, it is appropriate for what I call “policy research” where the aim is to learn the effects of variables on group behavior. For example, the group level data shows that wearing seat belts reduced fatalities in motor vehicle accidents. So based on this kind of “Casting Nets” data, people in policy making positions made it the law that everyone has to wear a seat belt while driving. Of course, wearing a seat belt only reduces fatalities on average; there are cases where people have died in a accident because they were wearing a seat belt (I know of two such cases personally). So there will always be some proportion of individuals in the group who are adversely affected by policies that have a beneficial effect on the group.Â
RM: Sometimes policies are only beneficial to the group if nearly the entire group abides by the policies that may have negative consequences for a very small subset of the individuals in the group. This is true, for example, of immunization which can protect everyone in the group, including those not immunized, if a large proportion of the group gets immunized. If a large proportion of the population refuses immunization due to the (extremely small) risks – which is happening with the small pox vaccine here in California – then the disease will remain active in the population and strike those who were unwilling or unable to get immunized.Â
RM: I do think all these group level problems will be solved once we understand the mechanisms that underlie the individual deviations. But until then, to the extent that we care about making things better at the group level, we’ll have to reply largely on the “Casting Nets” type of data, I think.Â
BestÂ
Rick
Â
Â
This problem isnât news for those of us in PCT it is one that Bill Powers railed against for deccades and is the thesis of Phil Runkelâs (1990) book, Casting Nets and Testing Specimens. Phil argued that large-N, group-based studies, with their random assignment of participants to treatments, statistical averaging, and inferential statistical analyses to estimate the reliability of any average differences between groups, are useful as screening tools for identifying potentially important relationships (âcasting netsâ?), but when it comes down to treating individuals, one must test individuals individually (âtesting specimensâ?), for the very reason noted in the quoted paragraph. PCT methodology, of course, does exactly that, as when performing âthe test for the controlled variable.â?Â
Â
The Discover article traces the history of attempts to apply this âsingle-subjectâ? or âN of 1â? approach when evaluating the effects of medical treatments on individual patients. âOne way to correct for the gaps the gold standard leaves in our knowledge is the âN of 1â? trial, where the number of participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis for example, âI think drinking milk will make me feel sick. Am I right?â? After a rocky start, the N of 1 approach is finally coming into its own under the banner of personalized medicine.  âSingle-subject trials allow you to personalize the outcome, not just the treatment. . . . That ability to improve the quality of care in a truly personalized way explains why N of 1 trials are making a comeback.â?
Â
Itâs an approach thatâs been at the heart of PCT from the beginning. Itâs nice to see that the medical world may finally be catching up.
Â
Bruce
–
Richard S. Marken, Ph.D.
Author of  Doing Research on Purpose.Â
Now available from Amazon or Barnes & Noble
I sent this post this morning but have no indication that it actually appeared and therefore am resending it. My apology if it’s a duplicate.
Bruce
Â
From: Bruce Abbott [mailto:bbabbott@frontier.com] Sent: Saturday, October 11, 2014 9:12 AM To: CSGnet Subject: Casting Nets and Testing Specimens
Â
[From Bruce Abbott (2014.10.11.0910 EDT)]
Â
November’s Discover magazine contains an article entitled “Singled Out� that points out a problem with the “gold standard� in medical research, the randomized control clinical trial used to evaluate the efficacy of drugs and other therapeutic measures. According to the article, “ . . . even the gold standard isn’t perfect. The controlled clinical trial is really about averages, and averages don’t necessarily tell you what will happen to an individual. Such a trial might tell you that, statistically speaking, milk isn’t good for Crohn’s patients. But within that sample, there might be people who didn’t have any problems drinking milk, and people whose symptoms even got better while drinking it.�
Â
This problem isn’t news for those of us in PCT – it is one that Billl Powers railed against for decades and is the thesis of Phil Runkel’s (1990) book, Casting Nets and Testing Specimens. Phil argued that large-N, group-based studies, with their random assignment of participants to treatments, statistical averaging, and inferential statistical analyses to estimate the reliability of any average differences between groups, are useful as screening tools for identifying potentially important relationships (“casting netsâ€?), but when it comes down to treating individuals, one must test individuals individually (“testing specimensâ€?), for the very reason noted in the quoted paragraph. PCT methodology, of course, does exactly that, as when performing “the test for the controlled variable.â€?Â
Â
The Discover article traces the history of attempts to apply this “single-subjectâ€? or “N of 1â€? approach when evaluating the effects of medical treatments on individual patients. “One way to correct for the gaps the gold standard leaves in our knowledge is the “N of 1â€? trial, where the number of participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis – for example, “I think drinking milk will make me feel ssick. Am I right?â€? After a rocky start, the N of 1 approach is finally coming into its own under the banner of personalized medicine.  “Single-subject trials allow you to personalize the outcome, not just the treatment. . . . That ability to improve the quality of care in a truly personalized way explains why N of 1 trials are making a comeback.â€?
Â
It’s an approach that’s been at the heart of PCT from the beginning. It’s nice to see that the medical world may finally be catching up.
I sent this post this morning but have no indication that it actually appeared and therefore am resending it. My apology if it’s a duplicate.
Bruce
Â
From: Bruce Abbott [mailto:bbabbott@frontier.com] Sent: Saturday, October 11, 2014 9:12 AM To: CSGnet Subject: Casting Nets and Testing Specimens
Â
[From Bruce Abbott (2014.10.11.0910 EDT)]
Â
November’s Discover magazine contains an article entitled “Singled Out� that points out a problem with the “gold standard� in medical research, the randomized control clinical trial used to evaluate the efficacy of drugs and other therapeutic measures. According to the article, “ . . . even the gold standard isn’t perfect. The controlled clinical trial is really about averages, and averages don’t necessarily tell you what will happen to an individual. Such a trial might tell you that, statistically speaking, milk isn’t good for Crohn’s patients. But within that sample, there might be people who didn’t have any problems drinking milk, and people whose symptoms even got better while drinking it.�
Â
This problem isn’t news for those of us in PCT – iit is one that Bill Powers railed against for decades and is the thesis of Phil Runkel’s (1990) book, Casting Nets and Testing Specimens. Phil argued that large-N, group-based studies, with their random assignment of participants to treatments, statistical averaging, and inferential statistical analyses to estimate the reliability of any average differences between groups, are useful as screening tools for identifying potentially important relationships (“casting netsâ€?), but when it comes down to treating individuals, one must test individuals individually (“testing specimensâ€?), for the very reason noted in the quoted paragraph. PCT methodology, of course, does exactly that, as when performing “the test for the controlled variable.â€?Â
Â
The Discover article traces the history of attempts to apply this “single-subjectâ€? or “N of 1â€? approach when evaluating the effects of medical treatments on individual patients. “One way to correct for the gaps the gold standard leaves in our knowledge is the “N of 1â€? trial, where the number of participants (N) is one instead of hundreds or thousands of volunteers. That one person works with a doctor to test a narrow hypothesis – for example, “I think drinking milk will make me feel sicck. Am I right?â€? After a rocky start, the N of 1 approach is finally coming into its own under the banner of personalized medicine.  “Single-subject trials allow you to personalize the outcome, not just the treatment. . . . That ability to improve the quality of care in a truly personalized way explains why N of 1 trials are making a comeback.â€?
Â
It’s an approach that’s been at the heart of PCT from the beginning. It’s nice to see that the medical world may finally be catching up.