Glossary entry (derived from question below)
Spanish term or phrase:
la condición matemática más desfavorable en los resultados esperados
English translation:
the least favourable mathematical outcome for the expected results
Added to glossary by
Lesley Jackson (X)
Feb 23, 2011 19:23
13 yrs ago
Spanish term
la condición matemática más desfavorable en los resultados esperados
Spanish to English
Science
Mathematics & Statistics
Spain: research paper:
Se recogieron y analizaron cuestionarios válidos de 319 especialistas en psiquiatría y de 957 pacientes depresivos, muestras que permiten una alta precisión en la estimación de resultados (con un error aleatorio de ± 5,5% en la encuesta a psiquiatras, y de ± 3,2% en la encuesta a pacientes, para un intervalo de confianza [IC] del 95%, supuesta una selección aleatoria de los encuestados y la condición matemática más desfavorable en los resultados esperados: p = q = 0,5).
What is the correct turn of phrase for this? (not my field) In Google I'm seeing "favorable/unfavorable result conditions" but can't pin down the entire thing. Thanks!
Se recogieron y analizaron cuestionarios válidos de 319 especialistas en psiquiatría y de 957 pacientes depresivos, muestras que permiten una alta precisión en la estimación de resultados (con un error aleatorio de ± 5,5% en la encuesta a psiquiatras, y de ± 3,2% en la encuesta a pacientes, para un intervalo de confianza [IC] del 95%, supuesta una selección aleatoria de los encuestados y la condición matemática más desfavorable en los resultados esperados: p = q = 0,5).
What is the correct turn of phrase for this? (not my field) In Google I'm seeing "favorable/unfavorable result conditions" but can't pin down the entire thing. Thanks!
Proposed translations
(English)
4 +1 | (assuming) the least favourable situation (that in which p=q=0.5.) for the expected results | DLyons |
Proposed translations
+1
7 hrs
Selected
(assuming) the least favourable situation (that in which p=q=0.5.) for the expected results
Discrimination is weakest where success probability = failure probability = 1/2.
--------------------------------------------------
Note added at 7 hrs (2011-02-24 02:27:26 GMT)
--------------------------------------------------
Also P.88 of
http://books.google.com/books?id=R-8d3xZLXGYC&printsec=front...
--------------------------------------------------
Note added at 23 hrs (2011-02-24 19:20:58 GMT)
--------------------------------------------------
Suppose, just for example, that depression comes in just categories "Mild", Moderate" and "Severe". A sample of 957 patients is taken to estimate the proportion of "Severe" cases in the population - suppose we find 189 "Severe" cases in the sample. That suggests about 20% of the population has severe depression (actually 19.7% = 189/957) . But that was only a sample, so the true value in the population may well be a bit more or less.
The question is "how much more or less?" and the study says ± 3.2%. Or to put it another way, we're 95% sure the true value is in the range [19.7-3.2, 19.7+3.2] = [16.6, 22.9]. This 95% is the "valor de significación".
What p is, is the estimated probability of having severe depression, i.e 0.197. And q is the estimated probability of NOT having severe depression i.e 1-0.197 = 0.803.
Finally, the ± 3.2% has been worked out based on the worst case scenario of p = q = 0.5 (logical if you think about it, it's hardest to discriminate if it's equally likely a person selected has or has not sever depression). So what they are saying is that 3.2 is conservative and, in practise, the range of [16.6, 22.9] can probably be reduced e.g. just to name a figure maybe it's really more like [18.0, 21.5] (still with 19.7 bang in the middle!)
Here endedth the lecture :-)
--------------------------------------------------
Note added at 7 hrs (2011-02-24 02:27:26 GMT)
--------------------------------------------------
Also P.88 of
http://books.google.com/books?id=R-8d3xZLXGYC&printsec=front...
--------------------------------------------------
Note added at 23 hrs (2011-02-24 19:20:58 GMT)
--------------------------------------------------
Suppose, just for example, that depression comes in just categories "Mild", Moderate" and "Severe". A sample of 957 patients is taken to estimate the proportion of "Severe" cases in the population - suppose we find 189 "Severe" cases in the sample. That suggests about 20% of the population has severe depression (actually 19.7% = 189/957) . But that was only a sample, so the true value in the population may well be a bit more or less.
The question is "how much more or less?" and the study says ± 3.2%. Or to put it another way, we're 95% sure the true value is in the range [19.7-3.2, 19.7+3.2] = [16.6, 22.9]. This 95% is the "valor de significación".
What p is, is the estimated probability of having severe depression, i.e 0.197. And q is the estimated probability of NOT having severe depression i.e 1-0.197 = 0.803.
Finally, the ± 3.2% has been worked out based on the worst case scenario of p = q = 0.5 (logical if you think about it, it's hardest to discriminate if it's equally likely a person selected has or has not sever depression). So what they are saying is that 3.2 is conservative and, in practise, the range of [16.6, 22.9] can probably be reduced e.g. just to name a figure maybe it's really more like [18.0, 21.5] (still with 19.7 bang in the middle!)
Here endedth the lecture :-)
Note from asker:
Thanks, DLyons. This being your specialty field, may I just ask you about the "p" here... is it the "valor de significación" P? or is it for "probability" or something else? |
I mean, I think you said as much in your good explanation... just want to confirm. Thanks. |
Peer comment(s):
agree |
Lafayette Eaton
: Good explanation Dlyons, this is a long-winded explanation of the confidence intervals of their sample.
1 hr
|
Thanks Lafayette.
|
4 KudoZ points awarded for this answer.
Comment: "With explanation way beyond the call of duty... thanks! "
Discussion