-

SCORING INSTRUCTIONS FOR THE MULTIPLE SCLEROSIS FUNCTIONAL COMPOSITE
:
The Z-score is a standardized number representing how close a test result is to the mean of a standard or reference population to which the result is compared. The Z-score is expressed in units of standard deviation. The Z-score is obtained by subtracting the mean of the reference population from the test result and then dividing by the standard deviation of the reference population. Because the Z-score is a relative measure indicating how many standard deviation units the current observation is from the mean of the reference population, the units are the same irrespective of the underlying measurement scale.
:
As an example, suppose we have five patients with an average time on the two trials of the 25-foot timed walk (TW) with times of 20, 25, 30, 35, and 40 seconds. The mean time in this group of patients is quite slow at 30 seconds, and the standard deviation is 7.906 seconds. To create a Z-score for the TW using these patients as the standard population, we subtract the overall average from each score and divide by the standard deviation of the population. Thus:
TABLE 4
Creating Z-scores Using The Test Patients To Standardize Scores
Patient Formula For Z-Score, TW Z-ScoreTW
#1 (20 – 30) / 7. 906 –1.27
#2 (25 – 30) / 7. 906 –0.63
#3 (30 – 30) / 7. 906 0.00
#4 (35 – 30) / 7. 906 0.63
#5 (40 – 30) / 7. 906 1.27
Patient #1 is 1.27 std dev units better than the mean, patient #2 is 0.63 std dev units better than the mean, patient #3 is 0 std dev units from the mean, patient #4 is 0.63 std dev units worse than the mean, and patient #5 is 1.27 std dev units worse than the mean.
:
The recommended scoring rule for clinical trials is to use the baseline average for statistical comparisons of the treatment and comparison groups. When comparing among studies, using the Task Force database values for showing results relative to an external standard is appropriate. These recommendations reflect the fact that many studies use a restricted range of subjects, who can yield different standard deviations among the components. Different standard deviations cause the three components to be weighted slightly different from study to study. An advantage in a clinical trial of using the average of all baseline measures is that the means and standard deviations of the component Z-scores will be close to zero and 1 respectively, making comparison of the treatments easier for assessing baseline comparability. Thus, in a clinical trial, it seems relevant to weight within the context of all the baseline data. However, until more experience is gained with the MSFC, it also seems reasonable to use the Task Force Standards to enable a single anchor comparison to describe the MSFC and the study populations. The task force standards were derived without adjustments for the learning effects on the component measures of the MSFC, which are now well documented. Thus, as more experience is gained with the MSFC and more data are accumulated, the task force standards should be updated.
(Link not provided, in order to maintain confidentiality)

statistical comparisons of the treatment => comparaisons statistiques du traitement

, it also seems reasonable to use the Task Force Standards to enable a single anchor comparison to describe the MSFC and the study populations. =>
, il semble également raisonnable d'utiliser les normes du groupe de travail pour permettre une comparaison à un seul ancrage pour décrire la MSFC et les populations étudiées.

Elle permet ainsi d’aborder non seulement la structure des éléments objectivés mais également l'ancrage de ces éléments dans les groupes sociaux.
:
Le cas des populations homogènes est concerné par la variabilité interindividuelle.
:
Les scores factoriels sont standardisés (moyenne = 0, écart-type = 1), et ce quelle que soit l’importance statistique du facteur concerné.
https://books.google.de/books?id=PWJBDQAAQBAJ&pg=PA171&lpg=P...

- 6 sujets avec lésion droite : 3 femmes et 3 homme, de 50 à 64 ans, leur moyenne d'âge étant de 57 ans (écart-type = 5,5). Ils n'ont pas de troubles du langage
:
- L’indice d’ancrage : Pour obtenir cet indice il nous a fallu d’abord obtenir l’ancrage des extrémités du continuum. Celui-ci est donné par deux mesures, l’une correspondant à la moyenne des valeurs des points situés à gauche de la pente et l’autre par celle des points situés à droite. … Afin de pouvoir comparer ces valeurs d'ancrage entre les différentes populations étudiées, un calcul supplémentaire a été introduit : le calcul de l’indice d’ancrage.
https://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd...