SPSSX Discussion

N = 1 Question

Classic

List

Threaded

5 messages Options

Salbod

N = 1 Question

Dear Friends,

What statistics are used in N=1 designs? A student just
asked me how to set up a data file for a single case study. He has no idea
about the statistics that will be used.

I was told that a single client was administered 8 measures (e.g., anxiety,
stress) at 6 different time periods during which time they were given a
treatment of some kind.

I've been spoiled working with large sample statistics and I'm not that
knowledgeable of small sample statistics (N=1). Can someone point me in the
right direction. Am I looking at a type of time series analysis here?

Any references or suggestions will be greatly appreciated.

Your friend, Stephen Salbod, Pace University, New York CIty

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Larry Burriss

Re: N = 1 Question

Stephen,

For nominal data you can use the binomial test or the one sample chi-square.
For ordinal data the Kolmogorov-Smirnov one-sample test and the one-sample
runs test will work.

NPAR TESTS BINOMIAL (specifications)/
CHISQUARE (specifications)/
K-S (specifications)/
RUNS (specifications).

/ Larry /

LARRY L. BURRISS, Ph.D., J.D.
Professor, School of Journalism
Middle Tennessee State University
Murfreesboro, TN 37132

615-898-2983
[hidden email]
http://www.mtsu.edu/~lburriss

***** This page made of 100% recycled electrons *****

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> Stephen Salbod
> Sent: Tuesday, December 04, 2007 2:59 PM
> To: [hidden email]
> Subject: N = 1 Question
>
> Dear Friends,
>
> What statistics are used in N=1 designs? A student just
> asked me how to set up a data file for a single case study. He has no idea
> about the statistics that will be used.
>
> I was told that a single client was administered 8 measures (e.g.,

anxiety,
> stress) at 6 different time periods during which time they were given a
> treatment of some kind.
>
> I've been spoiled working with large sample statistics and I'm not that
> knowledgeable of small sample statistics (N=1). Can someone point me in
the

> right direction. Am I looking at a type of time series analysis here?
>
>
>
> Any references or suggestions will be greatly appreciated.
>
>
>
> Your friend, Stephen Salbod, Pace University, New York CIty
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Arthur Kramer

Re: N = 1 Question

In reply to this post by Salbod

Years ago I obtained a book by Alan Kazdin on single case research designs.
You might want to see if it's still available.

Arthur Kramer, Ph.D.
Director of Institutional Research
New Jersey City University

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Stephen Salbod
Sent: Tuesday, December 04, 2007 3:59 PM
To: [hidden email]
Subject: N = 1 Question

Dear Friends,

What statistics are used in N=1 designs? A student just
asked me how to set up a data file for a single case study. He has no idea
about the statistics that will be used.

I was told that a single client was administered 8 measures (e.g., anxiety,
stress) at 6 different time periods during which time they were given a
treatment of some kind.

I've been spoiled working with large sample statistics and I'm not that
knowledgeable of small sample statistics (N=1). Can someone point me in the
right direction. Am I looking at a type of time series analysis here?

Any references or suggestions will be greatly appreciated.

Your friend, Stephen Salbod, Pace University, New York CIty

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Julius Sim

Re: N = 1 Question

In reply to this post by Salbod

Hi,

Yes, I think the Kazdin book is still in print - details of it, and some
others, below.

By 'six different time periods', do you mean only six observations (on
each outcome) in total? Or do you mean multiple serial observations in
each of six phases? If the latter, there is quite a lot of scope
statistically, including some simple time series analyses such as the C
statistic. If the former, I would have thought statistical (and indeed
visual) analysis would be a bit of a non-starter with so few datapoints.

Conventional tests for differences, like t tests and non-parametric
counterparts, are hard to use on n=1 data owing to the independence
assumption. If they are used, you need the independent t test (or Wilcoxon
rank sum) between phases, not the paired t test (or Wilcoxon signed
ranks), despite the apparent related nature of the data. But probably best
avoided if methods more suited to time series data are available. Not sure
how easily these are in SPSS, though - normally done it by hand!

Julius Sim

Barlow D H, Hersen M (1984) Single Case Experimental Designs: Strategies
for Studying Behavior Change, 2nd Edn. Pergamon Press, New York

Franklin R D, Allison D B, Gorman B S (1996) Design and Analysis of
Single-Case Research. Lawrence Erlbaum, Mahwah

Kazdin AE (1984) Single Case Research Designs: Methods for Clinical and
Applied Settings. Oxford University Press, New York

Kratochwill T R, Levin J R (1992) Single-Case Research Design and
Analysis: New Directions for Psychology and Education. Lawrence Erlbaum
Associates, Hillsdale

Ottenbacher K J (1986) Evaluating Clinical Change: Strategies for
Occupational and Physical Therapists. Williams & Wilkins, Baltimore

> Years ago I obtained a book by Alan Kazdin on single case research
> designs.
> You might want to see if it's still available.
>
> Arthur Kramer, Ph.D.
> Director of Institutional Research
> New Jersey City University
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> Stephen Salbod
> Sent: Tuesday, December 04, 2007 3:59 PM
> To: [hidden email]
> Subject: N = 1 Question
>
> Dear Friends,
>
> What statistics are used in N=1 designs? A student just
> asked me how to set up a data file for a single case study. He has no idea
> about the statistics that will be used.
>
> I was told that a single client was administered 8 measures (e.g.,
> anxiety,
> stress) at 6 different time periods during which time they were given a
> treatment of some kind.
>
> I've been spoiled working with large sample statistics and I'm not that
> knowledgeable of small sample statistics (N=1). Can someone point me in
> the
> right direction. Am I looking at a type of time series analysis here?
>
>
>
> Any references or suggestions will be greatly appreciated.
>
>
>
> Your friend, Stephen Salbod, Pace University, New York CIty
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>

Swank, Paul R

Re: N = 1 Question

In reply to this post by Larry Burriss

Doesn't the binomial test assume independence? If the data were binary
it sounds like perhaps some kind of runs test, perhaps the
Wald-Wolfowitz?

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Larry L. Burriss
Sent: Tuesday, December 04, 2007 4:06 PM
To: [hidden email]
Subject: Re: N = 1 Question

Stephen,

For nominal data you can use the binomial test or the one sample
chi-square.
For ordinal data the Kolmogorov-Smirnov one-sample test and the
one-sample
runs test will work.

NPAR TESTS BINOMIAL (specifications)/
CHISQUARE (specifications)/
K-S (specifications)/
RUNS (specifications).

/ Larry /

LARRY L. BURRISS, Ph.D., J.D.
Professor, School of Journalism
Middle Tennessee State University
Murfreesboro, TN 37132

615-898-2983
[hidden email]
http://www.mtsu.edu/~lburriss

***** This page made of 100% recycled electrons *****

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of
> Stephen Salbod
> Sent: Tuesday, December 04, 2007 2:59 PM
> To: [hidden email]
> Subject: N = 1 Question
>
> Dear Friends,
>
> What statistics are used in N=1 designs? A student
just
> asked me how to set up a data file for a single case study. He has no
idea
> about the statistics that will be used.
>
> I was told that a single client was administered 8 measures (e.g.,
anxiety,
> stress) at 6 different time periods during which time they were given
a
> treatment of some kind.
>
> I've been spoiled working with large sample statistics and I'm not
that
> knowledgeable of small sample statistics (N=1). Can someone point me
in
the

the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD