applied statistics
Independent samples: Tsinghua girls vs HKUST country girl
Behind small samples of normal authentication method:
The large sample sample variance that it is the population variance, and t -independent distribution.
Prerequisites: small sample two independent samples, and are normal population.
Standard deviation is the same, the degree of freedom direct addition, since the same shape.
Standard deviation is not the same, the image can not be transferred directly into a distribution, it is again adjusted weight:
Above, different standard deviation is calculated only on the different degrees of freedom.
Under normal circumstances are not the same as assuming that the variance
Pairing is generally non-independent nature in general, so in SPSS appears as the pair :
Two sets of evaluation criteria, different standards on the same subject of subjective difference is relatively small.
Large gap between the bodies on the use of experimental pair gap, experiment on a small body with Independent , because the original sample size is 2N , pair after the sample size is n- .
Than two normal population variance
Experimental Capacity Calculation:
Variance is known (it has that value, it does not estimate a) confidence level, sample size, margin of error is obtained,
When a variable rate, if the variance is unknown but want to known estimation variance equation, a conservative means may be employed, i.e., with variance represents [pi], [pi] is considered as 0.5 . Wherein, n- take the upper bound value, the formula for the relationship.