Need to calculate percentile score
I need help from one of you guru's. I have 8 statements that i ask, that are to be rated from 1 to 5 each. so, the lowest score is 8 and the highest could be 40. i also have a profile sheet, for reference, with the raw scores and %ile scores. for example, on the afroementioned scale, if the raw score is 17, the %ile score is 37. if the raw is 20, the %ile is 65. if the raw is 32, the %ile is 98, etc. OK, i'm no statistics wiz, and i've been out of grad school for a long time. i can figure the SD, z-score, Mean, or Median. i desperately need someone to give me a formula to calucate the %ile score from the information i have. i need this because i must plug this formula into an interpretative program i'm writing in access.
PLEASE HELP!
If your profile sheet gives every possible raw score and its corresponding percentile score, you can put them into Excel, create an X-Y scatter plot, and also put a trend line on the points (you can experiment with linear, polynomial, etc. to see which fits best). Excel will also give you the formula for this line.
y = 83.29Ln(x) - 186.4
this formula makes no sense to me. if "y" is the %ile score and "x" is the raw score, what is the value of Ln????
Ln stands for "natural logarithm"
Would you mind posting the raw score / percentile scores here? I can put it into Excel or a stats package and see what I get.....
Thanks so much for your help. I really do appreciate it. I did make sense out of the formula, but it is always just off a bit.
8 <35
9 <35
10 <35
11 <35
12 <35
13 <35
14 <35
15 <35
16 <35
17 37
18 47
19 56
20 65
21 72
22 77
23 82
24 86
25 88
26 90
27 92
28 93
29 95
30 96
31 97
32 98
33 99
34 99
35 >99
36 >99
37 >99
38 >99
39 >99
40 >99
Try using a polynomial, and increase the "order" to 6. You may also need to omit the more extreme points to get the best fit.....the inherent problem is that for the low and high points they give you "<" and ">" signs, which make formulations difficult.
This looks very promising, but, were a bit stupid here and we don't understand what the "E" stands for or what to do with it.
Let's say Excel gies you the equation:
y = 2E-05x6 - 0.0026x5 + 0.1834x4 - 6.7739x3 + 137.32x2 - 1437.5x + 6075.2
The "E" represents scientific notation, so 2E-05 is the same as 2.0 * 10^-5 or 0.00002
what represents the raw score in the notation you sent me. is the "x" scientific notation in the -.0026x5, or does it = .00265x10 to the fifth power???
it would be 0.00002 * x^6
plug the raw score into x, raise it to the 6th power, then multiply by 0.00002
thanks so much. i think this will work. sure do appreciate your patience with such a novice!
Posting Permissions
|
|
Reply With Quote