Please see the event/conference listing below to be advertised. Thank you
CONFERENCE: ANNOUNCEMENT AND CALL FOR ABSTRACTS
Weight stigma and health: whose problem is it?
University of Birmingham
Thursday 16th May, 2013
You are invited to attend a multidisciplinary conference focussing on
the subject of Weight Stigma and Health, featuring presentations from
both national and international keynote speakers. This conference will
be of interest to researchers, clinicians, and students from a wide
range of disciplines, including psychology, social sciences, medicine,
public health, allied health professions, sports and exercise science,
education, media studies, law, business and social policy.
We are also accepting abstracts for oral and/or poster presentations
on subjects related to weight-associated stigma, including, but not
limited to, weight stigma in health professionals, impact of weight
stigma on health and wellbeing, stigma in society, public policy
implications, legal considerations pertaining to weight-based
discrimination, and interventions to reduce weight stigma.
Postgraduate students are especially encouraged to submit their work.
Abstract submission instructions and the submission form can be found
on the website at http://stigmaconference.com/abstract-submission.
Please indicate if you would also like to be considered to deliver an
oral presentation. Abstract submission deadline is 7th April 2013.
The conference may also be of interest to second and third-year
undergraduate and masters students, who are encouraged to attend to
explore future research opportunities in the presence of staff from
various schools, disciplines, and locations around the UK and
overseas.
Details of keynote speakers and further conference information can be
obtained from www.stigmaconference.com.
Please contact Angela Meadows or Sara Tookey, Doctoral Researchers,
School of Psychology for further information:
[log in to unmask]
Sara A. Tookey
PhD Researcher
Frankland Building 309A
School of Psychology
University of Birmingham
Edgbaston
B15 2TT
On Wed, Mar 6, 2013 at 12:04 AM, PSYCH-POSTGRADS automatic digest
system <[log in to unmask]> wrote:
> There are 9 messages totaling 1062 lines in this issue.
>
> Topics of the day:
>
> 1. What does the t stand for in t-distribution? (4)
> 2. Reporting Cohen's d or Pearson's r effect sizes from a within-subjects
> design (2)
> 3. Within-subjects correlations (2)
> 4. SEM'n'R: PLS Path Modeling online education
>
> ----------------------------------------------------------------------
>
> Date: Tue, 5 Mar 2013 12:36:01 +0000
> From: Nina Bay <[log in to unmask]>
> Subject: What does the t stand for in t-distribution?
>
> Hi guys,
>
>
>
> I know that this is going to sound like a very basic question but what does
> the t stand for in t-distribution.
>
>
>
> I know that the t distributions were discovered by William S. Gosset in
> 1908 and that he was a statistician employed by the Guinness brewing
> company which had stipulated that he would not publish under his own name.
> He therefore wrote under the pen name ``Student.''
>
>
>
> But what does the t stand for?
>
>
>
> Many thanks
>
>
>
> Nina
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 12:40:46 +0000
> From: "STOCKER R." <[log in to unmask]>
> Subject: Re: What does the t stand for in t-distribution?
>
> Hi Nina,
>
> It stands for nothing, apparently! - it's just a notation, the letter t chosen arbitrarily... See here: http://www.isixsigma.com/topic/what-does-the-t-in-t-test-stand-for/
>
> Rachel
>
> From: Research of postgraduate psychologists. [mailto:[log in to unmask]] On Behalf Of Nina Bay
> Sent: 05 March 2013 12:36
> To: [log in to unmask]
> Subject: What does the t stand for in t-distribution?
>
>
> Hi guys,
>
>
>
> I know that this is going to sound like a very basic question but what does the t stand for in t-distribution.
>
>
>
> I know that the t distributions were discovered by William S. Gosset in 1908 and that he was a statistician employed by the Guinness brewing company which had stipulated that he would not publish under his own name. He therefore wrote under the pen name ``Student.''
>
>
>
> But what does the t stand for?
>
>
>
> Many thanks
>
>
>
> Nina
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 12:42:51 +0000
> From: Ben Haysom-Newport <[log in to unmask]>
> Subject: Re: What does the t stand for in t-distribution?
>
> I dispute that.
>
>
> From
> http://en.wikipedia.org/wiki/Talk%3AStudent's_t-distribution#what_does_.22t.22_stand_for.3F
>
> Well, basically, *t* stands for the domain of the probability
> distribution: [image:
> {-\infty} < t < \infty] ... much in the same way *z* stands for values of
> the domain of the [standardized] normal CDF and PDF, or [image: {\chi
> }^2] stands
> for domain values for the Chi-squared distributions.
>
> Best Wishes
> Ben Haysom-Newport MSc, BSc (Hons), MBPsS
>
>
> On 5 March 2013 12:36, Nina Bay <[log in to unmask]> wrote:
>
>> Hi guys,
>>
>>
>>
>> I know that this is going to sound like a very basic question but what
>> does the t stand for in t-distribution.
>>
>>
>>
>> I know that the t distributions were discovered by William S. Gosset in
>> 1908 and that he was a statistician employed by the Guinness brewing
>> company which had stipulated that he would not publish under his own name.
>> He therefore wrote under the pen name ``Student.''
>>
>>
>>
>> But what does the t stand for?
>>
>>
>>
>> Many thanks
>>
>>
>>
>> Nina
>>
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 07:10:49 -0800
> From: Jeremy Miles <[log in to unmask]>
> Subject: Re: What does the t stand for in t-distribution?
>
> It doesn't stand for something in the way that F stands for Fisher though.
> (It was named this by Snedecor, in older books it's sometimes called
> Snedecor's F).
>
> In stats, people often just choose a letter arbitrarily. That's why alpha
> is used for so many things (Cronbach intended alpha to be the first of a
> set of measures of reliability, the second was to be beta). Alpha is also a
> matrix of intercepts in a structural equation model, the p-value cutoff and
> the intercept in a regression model. Beta is the propability of a type II
> error, and the regression slope coefficients.
>
> Chi-squared is called chi-squared, because it looked (to Pearson) like a
> distribution he'd previously called chi. If you squared it.
>
> Jeremy
>
>
>
> On 5 March 2013 04:42, Ben Haysom-Newport <[log in to unmask]> wrote:
>
>>
>> I dispute that.
>>
>>
>> From
>> http://en.wikipedia.org/wiki/Talk%3AStudent's_t-distribution#what_does_.22t.22_stand_for.3F
>>
>> Well, basically, *t* stands for the domain of the probability
>> distribution: [image: {-\infty} < t < \infty] ... much in the same way *z* stands
>> for values of the domain of the [standardized] normal CDF and PDF, or [image:
>> {\chi }^2] stands for domain values for the Chi-squared distributions.
>>
>> Best Wishes
>> Ben Haysom-Newport MSc, BSc (Hons), MBPsS
>>
>>
>> On 5 March 2013 12:36, Nina Bay <[log in to unmask]> wrote:
>>
>>> Hi guys,
>>>
>>>
>>>
>>> I know that this is going to sound like a very basic question but what
>>> does the t stand for in t-distribution.
>>>
>>>
>>>
>>> I know that the t distributions were discovered by William S. Gosset in
>>> 1908 and that he was a statistician employed by the Guinness brewing
>>> company which had stipulated that he would not publish under his own name.
>>> He therefore wrote under the pen name ``Student.''
>>>
>>>
>>>
>>> But what does the t stand for?
>>>
>>>
>>>
>>> Many thanks
>>>
>>>
>>>
>>> Nina
>>>
>>
>>
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 15:50:25 +0000
> From: Michael Puniskis <[log in to unmask]>
> Subject: Reporting Cohen's d or Pearson's r effect sizes from a within-subjects design
>
> Hi Everyone
>
> I have a question about reporting effect size but am finding it difficult to get a clear answer from a lot of sources I have read online. Basically I was wondering if it is statistically correct to report either Cohen's d or Pearson's r effect sizes from a within-subjects design with data analysed through a one-tailed paired t-test.
>
> The data itself is collected through pre and post questionnaires for a social work intervention. In our evaluation reports we report the percentage change between pre- and post-test means, however I would also like to report this in a more scientific way through effect size if possible.
>
> In hours of online research I've only found very few sources to say this is possible (or at least to specifically say so), as the majority of information I find seems to say effect sizes can only be reported when you have a between-subjects design, for example with a treatment and control group.
>
> Any information would be greatly appreciated, thanks in advance.
>
> Mike
>
>
>
> ---------------------------------------------------------------------------
>
>
> Please note that Middlesex University's preferred way of receiving all correspondence is via email in line with our Environmental Policy. All incoming post to Middlesex University is opened and scanned by our digital document handler, CDS, and then emailed to the recipient.
>
> If you do not want your correspondence to Middlesex University processed in this way please email the recipient directly. Parcels, couriered items and recorded delivery items will not be opened or scanned by CDS. There are items which are "exceptions" which will be opened by CDS but will not be scanned a full list of these can be obtained by contacting the University.
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 16:12:23 +0000
> From: Juliet Wakefield <[log in to unmask]>
> Subject: Within-subjects correlations
>
> Hello,
>
> I have a questionnaire-based experimental study (N = 59) involving two conditions. After giving participants the experimental manipulation, we presented the dependent measure. This consisted of a list of 24 personality traits (some positive, some negative). For each trait, we asked the participant to rate on a 1-7 scale: i) how much they possess this characteristic; and ii) how desirable it is to possess this characteristic. We want to compare the relationship between perceived possession and perceived desirability in the two conditions.
>
> The author of the paper that we based the study on recommended that we run a within-subjects correlation, where for each participant we obtain the correlation between perceived possession and perceived desirability for each of the 24 traits, and then standardize these correlations before comparing them.
>
> Does anybody know how you run a within-subjects correlation in SPSS (and how you save the resultant correlations into the data-file)?
>
> Many thanks for your help,
>
> Juliet.
>
> Dr. Juliet Wakefield,
> Post-Doctoral Researcher,
> Room 4.33,
> School of Psychology,
> University of Dundee,
> Dundee DD1 4HN.
> Tel: 01382 384853
> Web: http://www.dundee.ac.uk/psychology/people/research/jrhwakefield/
> Project site: http://www.healthingroups.org<http://www.healthingroups.org/>
> Twitter: @healthingroups
> Facebook: http://facebook.com/healthingroups
>
>
> The University of Dundee is a registered Scottish Charity, No: SC015096
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 09:09:56 -0800
> From: Jeremy Miles <[log in to unmask]>
> Subject: Re: Within-subjects correlations
>
> If I've understood correctly, you have 24 pairs of measures per
> person, so your data could look like:
>
> Trait Person Possession Desirability
> 1 1 4 5
> 2 1 1 4
> 3 1 7 5
> 4 1 6 3
> ...
> 24 1 5 2
> ...
> 1.. 59 4 3
> ...
> 25 59 1 6
>
> If they don't, you'll be able to make them look like that with by
> using the transform -> reshape option.
>
> If that's right, use split file to estimate the correlation for each person.
>
> Then use the Output Management System (OMS), which lets you take
> output and turn it into data. You'll need to fiddle around to do that.
>
> Then run the standardization (which is in descriptives), with (I
> think) split file on again.
>
> SPSS isn't designed for weird data structures like this - it's
> sometimes easier to fiddle with the data in a program that is more
> flexible, I tend to use either R or Excel.
>
> Jeremy
>
>
> On 5 March 2013 08:12, Juliet Wakefield <[log in to unmask]> wrote:
>> Hello,
>>
>>
>>
>> I have a questionnaire-based experimental study (N = 59) involving two
>> conditions. After giving participants the experimental manipulation, we
>> presented the dependent measure. This consisted of a list of 24 personality
>> traits (some positive, some negative). For each trait, we asked the
>> participant to rate on a 1-7 scale: i) how much they possess this
>> characteristic; and ii) how desirable it is to possess this characteristic.
>> We want to compare the relationship between perceived possession and
>> perceived desirability in the two conditions.
>>
>>
>>
>> The author of the paper that we based the study on recommended that we run a
>> within-subjects correlation, where for each participant we obtain the
>> correlation between perceived possession and perceived desirability for each
>> of the 24 traits, and then standardize these correlations before comparing
>> them.
>>
>>
>>
>> Does anybody know how you run a within-subjects correlation in SPSS (and how
>> you save the resultant correlations into the data-file)?
>>
>>
>>
>> Many thanks for your help,
>>
>>
>>
>> Juliet.
>>
>>
>>
>> Dr. Juliet Wakefield,
>>
>> Post-Doctoral Researcher,
>>
>> Room 4.33,
>>
>> School of Psychology,
>>
>> University of Dundee,
>>
>> Dundee DD1 4HN.
>>
>> Tel: 01382 384853
>>
>> Web: http://www.dundee.ac.uk/psychology/people/research/jrhwakefield/
>>
>> Project site: http://www.healthingroups.org
>>
>> Twitter: @healthingroups
>>
>> Facebook: http://facebook.com/healthingroups
>>
>>
>>
>>
>> The University of Dundee is a registered Scottish Charity, No: SC015096
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 09:13:24 -0800
> From: Jeremy Miles <[log in to unmask]>
> Subject: Re: Reporting Cohen's d or Pearson's r effect sizes from a within-subjects design
>
> You can either calculate a Cohen's D in the usual way, or you can
> convert the F statistic to some sort of effect size (which is covered,
> I think, in Andy Field's "Discovering Statistics..." books).
>
> Also, while I'm here, it's very unlikely you should really be using a
> one tailed test. A one tailed test means a result in the opposite
> direction to that which you expect is theoretically and substantively
> meaningless, and the actions you take as a result will be the same as
> if you found no effect at all. If that's not true, when you use a
> one-tailed test, you're just lying about your p-value.
>
> (I know that sounds harsh, but I do a LOT of statistical reviews, and
> if you put a one tailed test in a paper I review, you will be told to
> revise it.)
>
> Jeremy
>
>
> On 5 March 2013 07:50, Michael Puniskis <[log in to unmask]> wrote:
>> Hi Everyone
>>
>> I have a question about reporting effect size but am finding it difficult to get a clear answer from a lot of sources I have read online. Basically I was wondering if it is statistically correct to report either Cohen's d or Pearson's r effect sizes from a within-subjects design with data analysed through a one-tailed paired t-test.
>>
>> The data itself is collected through pre and post questionnaires for a social work intervention. In our evaluation reports we report the percentage change between pre- and post-test means, however I would also like to report this in a more scientific way through effect size if possible.
>>
>> In hours of online research I've only found very few sources to say this is possible (or at least to specifically say so), as the majority of information I find seems to say effect sizes can only be reported when you have a between-subjects design, for example with a treatment and control group.
>>
>> Any information would be greatly appreciated, thanks in advance.
>>
>> Mike
>>
>>
>>
>> ---------------------------------------------------------------------------
>>
>>
>> Please note that Middlesex University's preferred way of receiving all correspondence is via email in line with our Environmental Policy. All incoming post to Middlesex University is opened and scanned by our digital document handler, CDS, and then emailed to the recipient.
>>
>> If you do not want your correspondence to Middlesex University processed in this way please email the recipient directly. Parcels, couriered items and recorded delivery items will not be opened or scanned by CDS. There are items which are "exceptions" which will be opened by CDS but will not be scanned a full list of these can be obtained by contacting the University.
>
> ------------------------------
>
> Date: Tue, 5 Mar 2013 13:46:33 -0500
> From: Geoffrey Hubona <[log in to unmask]>
> Subject: SEM'n'R: PLS Path Modeling online education
>
> 14-Course, 9-month online PLS and SEM program taught by International
> experts.
> $295-$345 early registration cost (before Apr 1) for entire program.
> Receive PLS courseware partial downloads upon registration.
> http://sem-n-r.com/Courses.html
>
> Questions to [log in to unmask]
>
> Geoff Hubona
>
> ------------------------------
>
> End of PSYCH-POSTGRADS Digest - 4 Mar 2013 to 5 Mar 2013 (#2013-52)
> *******************************************************************
|