where su is the expected standard error of prediction (uncertainty) for the new observation u and n is the number of training set observations t used to build the model.
Therefore it is important to calculate the expected standard error before committing resources to collecting the data.
Using results from Hanley [ 39], the expected standard error of kappa in a sample of 120 will be in between.07 (for kappa 0.3) and.05 (for kappa =.8).
In this form of study, the aim is to control the expected standard error of the estimate derived from the sample.
The sample size and the number of questions were varied systematically until the expected standard error associated with the utility estimates was 0.05 or smaller.
The composite information across all the ADAS-Cog items was used to estimate the expected standard error of measurement of different levels of cognitive impairment using the ADAS-CogIRT methodology.
In our study, the final sample size was based on the past experience; we use the expected standard error of 0.05 recommended by Sawtooth Software, so our results could be compared more easily with the results of other studies that use CBC.
The presence of adequate targeting of the items to the sample, the sample size required to perform a Rasch analysis should have included at least 200 observations in order to yield stable person and item estimates (±0.5 logit at 95% confidence level) and based on an expected standard error level of ±0.1 [ 71].
The expected standard errors/widths of the 95% confidence intervals for each prevalence estimate will be as follows (see Table 2): The minimum difference in prevalence rates between two pre-specified countries that can be detected with a power of 80% or 90% is 9.2% (from 34.6 to 25.4%, risk reduction 32.0%) or 9.4% (from 34.7 to 25.3%, risk reduction 36.2%).
We will introduce important concepts such as random variables, independence, Monte Carlo simulations, expected values, standard errors, and the Central Limit Theorem.
The accuracy evaluation is performed by comparing the expected relative standard errors (RSE) from the MC/AGQ FIM with RSE obtained in a simulation study with four different discrete data models (two binary, one count and one repeated time-to-event model) and three different estimation algorithms.
