If you are interested in assessment memory impairment as an important component of cognitive impairment, I would recommend to use California Verbal Learning Tests, second edition (CVLT-II). We found it useful in assessment memory impairment in patients with multiple sclerosis.

Dr. Igor Stepanov

My experience with the CANTAB is that it is a good measure of reaction time and processing speed. I think that the way it is done, with easy to more difficult decisions to be made, with redundancy in the reaction time measures over the course of the exam itself (if you do the regular one and not the mini-battery), are all nice. If you are interested in the impact that reaction time has on cognition, then the CANTAB may be helpful. However, it is a limited sample of cognition itself....as pointed out by Stepanov.

It depends a great deal on the subject group and the question. For example, Raven's advanced progressive matrices can be a useful measure of reasoning and abstraction skills in high-functioning adults, while we often use measures with a substantial executive and self-management component in our research work (for example, the selective reminding test, measures of divided attention, and computerized measure such as the attention networks test). For peak performance, tasks must be sufficiently challenging to involve motivation and persistence, and not have a clear ceiling effect.

It would help if you can be more clear on they types of cognitive abilities you hope to assess. Cognition covers many different constructs and measures for such constructs.

Agree with Alan... it depends on what cognitive skills you want to measure and whether you are looking for differences from the general population or pre-post training effects.

Thanks for the replies so far and for asking for clarification. We are looking to measure peak physical and cognitive performance of 18-30 year old males. We will have 2 groups, 1 a normal group and an elite sports sample group (rowers perhaps). There will be an 8 week training intervention and there will be a supplementation/ placebo utilised too. We will be trying to test these groups 4 times (pre, mid, post and follow up) and so with the proposed tests we are looking to control for practice effects and need sensitive enough tests that may show improvements on cognition post exercise and/or supplementation. I understand peak cognitive performance is broad to say the least but I think we will try and focus on executive functioning and visual memory/attention.Again some of you may have a better handle on this and therefore- Any suggestions would be greatly appreciated.

So young healthy adults, and repeated measures. This sounds like some of the work that we're doing with older adults. We chose the Selective Reminding Test (a complex measure with multiple forms), the Medical College of Georgia figures (again, with 4 forms), and we used the D-KEFS sorting test, using one set of cards at each time, which was sufficient to show an effect with n=40. Executive and attention testing is challenging in this kind of study, as so many tasks are sensitive to practice effects. It may make sense to develop some alternate test forms for attention measures, or see what is available and sufficiently challenging (such as the PASAT, n-back, or Brown-Peterson formats).

for Episodic memory: CVLT,/ Inhibtion: Hayling task +go no go test+ stroop Task/ Cognitive flexibility (executive function) TMT and Wisconsin card, you can use also, verbal fluency test. Good luck

Mark: OK, I still think there's room for more precision for the measures to reflect your hypotheses. What kinds of *functional* changes do you expect to see in your intervention group? Or not to see in your supplementation group? (e.g., peak performance for a rower might be quite different than for a golfer). If you can share the nature of your intervention(s) without feeling it's over-sharing, that might point to parallels in terms of underlying skills. To the extent you can refer to functional changes, without just measuring what you're training (of course you don't want to do that), you can do a cognitive task analysis of the functional activities, then select the core common skills that may be expected to change.

Sharpening aside - I would think in terms of tests not primarily designed to detect pathology that don't have a lot of practice effect potential over a short term:

- speed and efficiency of information-processing (so maybe something like Computerized Test of Information Processing to see if they can beat the average or their own scores after training; or paced tests that could be computerized to mtch their speed - like Posit's assessment/training measures that increase pacing based on individual performance. You shouldn't get training effects from this kind of measure if they are repeating it only once per 4 weeks.)

- working memory (dual n-back; perhaps determining their baseline level, then measuring change OR % success at a pre-determined level, but make sure that level is relatively challenging)

- Cognitive fluency (NAB category fluency; semantic fluency; design fluency)

- And perhaps consider a physiological measure like heart-rate variability as a measure of stress response - increased peak performance might be associated with better stress management/less emotional reactivity for some groups.

Your challenge is that you have young healthy adults and you're looking for fine distinctions in the upper ranges. Most tests designed to detect pathology (e.g., WCST) won't help you with this, since most people can do it (that's kind of the point) and there are limited distinctions in the upper performance ranges in the standardization groups, since that's not the point of the test. So if you can find tasks that have scores that can be compared with previous performance (e.g., fluency, RT) rather than seen as practice with specific materials (e.g., in my experience, once someone has done the WCST successfully once, they remember what they did the next time -- even a year later), this will work better for you.

Hope that helps Something.

From what I have been reading, above all else don't forget to include a throw-away or practice administration. The learning that will take place between the practice and and baseline (T1) administrations could be greater than any improvements you find between your time-points of interest (T1-4).

My dear Mark, talk about measuring the cognitive potential is very broad!!

First you have to assess cognitive functions which you want to "highlight". There are many tests that assess general cognition, but other tests assess specific functions. If your research is for clinical interest, make a battery of standardized tests and validated for your sample. Validation it is important point!

However, if you conduct a research and investigate a specific pathology, see what others neuropsychologists have used in their studies.

If interested, please send me more detailed in your question and if I can, I am pleased to help you.

Best regards,

Livia

If you are trying to measure peak cognitive performance in a normal population (e.g. perhaps, given your field of sport psychology, you are trying to get a baseline that will be sensitive to later concussive injury that may have subtle effects) then you want to tap performances that require integration of multiple brain functions, that do not have ceiling effects, and where raw scores can be meaningfully compared across repeated performances. Timed tests of complex functions (i.e. there is no ceiling on time), that have validated alternate forms would fit the bill, OR (here is a pet topic of mine) it would be good if the field of Neuropsychology included a strong research base on expected learning curves with repeated administration. We really need to stop dismissing "practice effects" as error variance and look at this as a meaningful, informative phenomena. Anway, back to your question: if this is related to clinical assessment, getting a baseline prior to potential injury, and then the measures are repeated after injury, the other question you'll have to address is how convenient the measures need to be. Some of the timed measures of integrated brain functions - e.g. the Category Test, the TPT - are time consuming. They are also likely influenced by practice effects. For a clinical question, however, if someone did at all worse on these after a second, post-injury administration, it would be highly suspicious for new (relative) impairment. A more convenient measure that, interestingly, does seem to tap widespread brain functioning, is good old "Finger tapping", adminstered according to the standardized instructions. See Prigatano's fascinating body of work on finger tapping. Hope to hear back from you with more details. Bes regards,

Sorry, I missed your reply to other questions, clarifying your design. Some of my suggestions are still relevant, however.

Hello Nanci,

the ideal is to measure the cognitive potential and overall performance of a subject in neuropsychological tasks is a comprehensive test (need not be a test for specific functions) and supplement with a battery of tests to assess specific functions. For instance: test for attention (selective, alternating, concentrated), Test for memory (visual semantics, immediate, work, long-term) executive (inhibitory control, fluency of thought) and other functions as more refined visual construction, visual perception, among others.

You can get information about which tests to use your battery in the book: Strauss E SE, Spreen O A compendium of neuropsychological tests: administration, norms, and commentary. Oxford University Press. 2006:1-1216.

I hope I have helped,

Kind regards,

Livia

Running memory span

thanks Karen, Nancy, Livia, James and Maryanne. This is really very helpful to me and has given me some real food for thought. As I want to test the participants on at least 4 occasions I am very wary of practice effects and sensitivity to change and would like to select the best tests to measure exactly what I am looking for. We will have a battery of physiological measures too including HRV, blood samples and muscle biopsys. A pertinent question now really for me is should I have one for example a WM test at one stage and another/ different WM test at the next time of testing and so on...?

Mark -- to the extent possible you want to use the same measure or the same measure with matched versions. If, for example, you use the N-back test at a fairly challenging level (e.g., at least 3 after an initial training run to give them the idea of the measure), you shouldn't get training effects if they do a couple of runs each time. You don't want any variation in skill performance to result from the tests having slightly (or widely) differing challenge levels. That's why I was thinking of measures like CTIP, N-back, fluency -- there aren't large practice effects with these.

Sorry.. I would insert me in such question. Our studies have explained that quantum mechanics has a role in perceptive as well as in cognitive functions. See my list of publications. In order to give proof of this we have used standard test at perceptive as well as cognitive level. However, the investigation by quantum mechanics enables us to identify indexes that may be of interst. Why don't use also such texts?.

Evaluation of Reaction Times?

Mark, re WM tests, a clear finding over the years re Neuropsychilogical tests is that tests designed to measure the same thing are not interchangeable. Different features of the stimulus materials and presentation method have unknown, sometimes large effects. Different "normative samples" have different results too. I agree w previous comments that you need alternate forms of a test using the same normative group.