Positive Predictive Value — ppv • ClinicoPath

Calculates the Positive Predictive Value (PPV) and False Discovery Rate (FDR) for research findings based on study characteristics. This tool helps researchers understand the probability that their significant findings are actually true, considering factors like prior probability, statistical power, and research biases.

Usage

ppv(percTrue = 50, alpha = 0.05, power = 0.8, percHack = 0)

Arguments

percTrue: The pre-study probability that the tested relationships are true. In exploratory research fields, this may be 10\ confirmatory research with strong theoretical basis, it may be higher.
alpha: The significance level (Type I error rate) used in the studies. Standard value is 0.05, but may be lower for multiple testing situations.
power: Statistical power of the studies to detect true effects. Well-designed studies typically have 80\ studies have much lower actual power.
percHack: Percentage of studies with questionable research practices (p-hacking, selective reporting, data dredging). This represents bias in the research process that increases false positive rates.

Value

A results object containing:

`results$confusion`					a table containing the true/false positives/negatives
`results$ppv`					a html
`results$dotPlot`					an image

Tables can be converted to data frames with asDF or as.data.frame. For example:

results$confusion$asDF

as.data.frame(results$confusion)

Examples

# Calculate PPV for a typical medical research scenario
ppv(
    percTrue = 10,    # 10\% of tested hypotheses are true
    alpha = 0.05,     # 5\% significance level
    power = 0.80,     # 80\% power
    percHack = 30     # 30\% of studies have some bias
)