DATA Subcommand (META CONTINUOUS command)
The META CONTINUOUS command calls the meta-analysis procedure with continuous outcomes and raw effect size data. The required DATA subcommand specifies the meta-analysis data.
- TREATMENT
- The required keyword specifies the data for the treatment group.
- N
- The required keyword specifies the sample size followed by a required variable name that is specified in parentheses. The value must be a single numeric variable.
- MEAN
- The required keyword specifies the sample means followed by a required
variable name that is specified in parentheses. The values must be a single numeric variable. The
value cannot be the same as the value specified for
CONTROL = MEAN(variable). - VAR
- The required keyword specifies the sample variance followed by a required variable name that is specified in parentheses. The value must be a single numeric variable. The keyword cannot be specified together with STD.
- STD
- The required keyword specifies the sample standard deviation followed by a required variable name that is specified in parentheses. The value must be a single numeric variable. The keyword cannot be specified together with VAR.
- CONTROL
- The required parameter specifies the data for the control group.
- N
- The required keyword specifies the sample size followed by a required variable name that is specified in parentheses. The value must be a single numeric variable.
- MEAN
- The required keyword specifies the sample means followed by a required
variable name that is specified in parentheses. The values must be a single numeric variable. The
value cannot be the same as the value specified for
TREATMENT = MEAN(variable). - VAR
- The required keyword specifies the sample variance followed by a required variable name that is specified in parentheses. The value must be a single numeric variable. The keyword cannot be specified together with STD.
- STD
- The required keyword specifies the sample standard deviation followed by a required variable name that is specified in parentheses. The value must be a single numeric variable. The keyword cannot be specified together with VAR.
- ESTYPE
- The optional keyword specifies the effect size type.
- COHEN_D
- Estimates the Cohen’s d. This is the default setting.
- COHEN_D(ADJUST)
- Estimates the Cohen’s d and its variance by using an alternative formula
divided by
2(Ntreatment + Ncontrol −2). - HEDGES_G
- Estimates the Hedges’ g.
- HEDGES_G(ADJUST)
- Estimates the Hedges’g and its variance by using an alternative formula
divided by
2(Ntreatment + Ncontrol −3.94). - GLASS_DELTA
- Estimates the Glass’s ∆ based on the control group.
- GLASS_DELTA(TREAT)
- Standardizes the Glass’s ∆ based on the standard deviation of the treatment group.
- MEAN_DIFF
- Estimates the mean difference by assuming the two population standard deviations are equal.
- MEAN_DIFF(UNEQUAL)
- Estimates the mean difference by assuming the two population standard deviations are not equal.
- ID
- The optional keyword specifies the study ID (usually a numeric variable). The variable that is specified for this parameter cannot be the same as the variable that is specified for STUDY.
- STUDY
- The optional keyword specifies the study name (usually a string variable). Either STUDY, or ID, or both must be specified to identify each primary study. The variable that is specified for this parameter cannot be the same as the variable that is specified for ID.