RoBSA (version 1.0.3)

prior_informed: Creates an informed prior distribution based on research

Description

prior_informed creates an informed prior distribution based on past research. The prior can be visualized by the plot function.

Usage

prior_informed(name, parameter = NULL, type = "smd")

Value

prior_informed returns an object of class 'prior'.

Arguments

name

name of the prior distribution. There are many options based on prior psychological or medical research. For psychology, the possible options are

"van Erp"

for an informed prior distribution for the heterogeneity parameter tau of meta-analytic effect size estimates based on standardized mean differences (van Erp et al. 2017),

"Oosterwijk"

for an informed prior distribution for the effect sizes expected in social psychology based on prior elicitation with dr. Oosterwijk (Gronau et al. 2017).

For medicine, the possible options are based on Bartoš et al. (2021) who developed empirical prior distributions for the effect size and heterogeneity parameters of the continuous standardized outcomes based on the Cochrane database of systematic reviews. Use "Cochrane" for a prior distribution based on the whole database or call print(prior_informed_medicine_names) to inspect the names of all 46 subfields and set the appropriate parameter and type.

parameter

parameter name describing what prior distribution is supposed to be produced in cases where the name corresponds to multiple prior distributions. Relevant only for the empirical medical prior distributions.

type

prior type describing what prior distribution is supposed to be produced in cases where the name and parameter correspond to multiple prior distributions. Relevant only for the empirical medical prior distributions.

Details

Further details can be found in erp2017estimates;textualBayesTools, gronau2017bayesian;textualBayesTools, and bartos2021bayesian;textualBayesTools.

References

See Also

prior(), prior_informed_medicine_names

Examples

# prior distribution representing expected effect sizes in social psychology
# based on prior elicitation with dr. Oosterwijk
p1 <- prior_informed("Oosterwijk")

# the prior distribution can be visualized using the plot function
# (see ?plot.prior for all options)
plot(p1)


# empirical prior distribution for the standardized mean differences from the oral health
# medical subfield based on meta-analytic effect size estimates from the
# Cochrane database of systematic reviews
p2 <- prior_informed("Oral Health", parameter = "effect", type = "smd")
print(p2)