Accessing results available when expoiting exchangeability

 

Required sample size

 

Usage

• findsize  operator; two operands, N followed by E, in parenthesis and separated by a comma. N is the name of an element or base. E is any valid equation. The result of the equation, should be a real number a satisfying .

This operator may be used only when an exchangeable adjustment of the type introduced in §16.2 is available. Thus, we assume that some adjustment has been performed. The findsize  operator is used to determine sample sizes to guarantee various degrees of resolution for both the collection as a whole and for its component elements. This operator may be used to duplicate the output shown when the FINDSIZE:  command is issued.

In the first instance, N should be the name of the collection adjusted, . The number returned is the sample size n such that . If the resolution a is unattainable, a value of is returned.

In the second instance, N should be the name of an element of the collection adjusted, . Then the number returned is the sample size n such that . If the resolution a is unattainable, a value of is returned.

An example of the use of the command is as follows, where Patients is assumed to be the collection currently being adjusted.

BD>print : (findsize (Patients,0.95))

 

Resolution for a given sample size

 

Usage

• resd  operator; two operands, N followed by E, in parenthesis and separated by a comma. N is the name of an element or base. E is any valid equation. The result of the equation when rounded should be a positive integer .

This operator may be used only with when an exchangeable adjustment of the type introduced in §16.2 is available. Otherwise the operator can be used with to return the current resolution. We assume that some adjustment has been performed. The resd  operator is used to calculate the resolution in N for a sample of size m. It works independently of the current sample size, n.

In the first instance, N should be the name of the collection adjusted, . The number returned is the resolution in B given by the adjustment by , . In the second instance, N should be the name of an element of the collection adjusted, . The number returned is the resolution in X given by the adjustment by , .  

Resolved variance for a given sample size

 

Usage

• rvard  operator; two operands, N followed by E, in parenthesis and separated by a comma. N is the name of an element or base. E is any valid equation. The result of the equation when rounded should be a positive integer .

This operator may be used only with when an exchangeable adjustment of the type introduced in §16.2 is available. Otherwise the operator can be used with to return the current resolved variance. We assume that some adjustment has been performed. The rvard  operator is used to calculate the resolved variance in N for a sample of size m. It works independently of the current sample size, n.

In the first instance, N should be the name of the collection adjusted, . The number returned is the resolved variance in B given by the adjustment by , . In the second instance, N should be the name of an element of the collection adjusted, . The number returned is the resolved variance in X given by the adjustment by , .  

Adjusted variance for a given sample size

 

Usage

• vard  operator; two operands, N followed by E, in parenthesis and separated by a comma. N is the name of an element or base. E is any valid equation. The result of the equation when rounded should be a positive integer .

This operator may be used only with when an exchangeable adjustment of the type introduced in §16.2 is available. Otherwise the operator can be used with to return the current adjusted variance. We assume that some adjustment has been performed. The vard  operator is used to calculate the adjusted variance in N for a sample of size m. It works independently of the current sample size, n.

In the first instance, N should be the name of the collection adjusted, . The number returned is the adjusted variance in B given by the adjustment by , . In the second instance, N should be the name of an element of the collection adjusted, . The number returned is the adjusted variance in X given by the adjustment by , .