Say that you know 70% of the material. When the number of questions is small, the standard deviation of your result is large (N-1…it’s used simply because of this reason). When the number of question is large, the standard deviation will decrease and your result will be closer to 70%. Therefore, a smaller number of questions makes it more like to score <65% than when the number of questions is larger. Could anyone do a t-stat for this one??? 
gz2nyc Wrote: ------------------------------------------------------- > “1 question to pass level II or 500?” > > If time is a pressuring factor and I’m limited to > the same amount of time, I’ll pick 1 question. That is why I said “If you have already studied and learned everything to the best of your ability you would pick 500 of course.” If you aren’t ready of course you pick 1. I would too. That would be my choice right now.
DblA Wrote: ------------------------------------------------------- > From CFAI c. 1997: > > The Board of Trustees of the ICFA is responsible > for establishing the pass-fail mark for each level > of the CFA Program. The initial pass mark is 70% > of the top 1% of papers. The final pass mark is > determined by the Trustees, who look for guidance > to psychometric consultants employing the modified > Angoff Method, a standard-setting method widely > accepted throughout the education testing field. > -------------------- > > This is 10 years old, but they said basically the > same thing last year. > They actually have qualified this statement quite alot. First of all there is no first cut at 70% of the top 1%. The MPS (minimum passing score) is determined by the modified Angoff method and then adjusted for soft factors if appropriate. I won’t bore you with the details of the Angoff method, but it basically consists of asking number of panelists to assess the percentage of minimally qulified Level II candidates who would get each question correct. The average response of each panelist is averaged with the other panelists to produce the minimum passing score. Ethics can break close calls (from a CFAI presentation). And the net result of the Angoff method just happens to be pretty close to the old way of doing things.