I copied and pasted these from Table 3 in your manuscript.
pvalues <- c(0.00032,
0.00023,
0.035,
0.023,
0.034,
0.00098,
0.0013,
0.127,
0.838,
0.475,
0.343,
0.00043,
0.109,
0.129,
0.165)
labels <- c("Competency/Knowledge base (Positively Mentioned)",
"Thoroughness (Positively Mentioned)",
"Thoroughness (Negatively Mentioned)",
"Temperament (Negatively Mentioned)",
"Cost-consciousness (Negatively Mentioned)",
"Interactions with staff (Positively Mentioned)",
"Interactions with staff (Negatively Mentioned)",
"Billing and insurance (Positively Mentioned)",
"Parking (Negatively Mentioned)",
"Black",
"Other",
"Knew Name of Provider (Yes, Reference = No)",
"Responded to comments (Yes, Reference = No)",
"Practice Type (Academic, Reference = Non-academic)",
"Constant")
The default method is the “Holm” method. According to the R documentation: “There seems no reason to use the unmodified Bonferroni correction because it is dominated by Holm’s method, which is also valid under arbitrary assumptions.”
adj.pvalues <- p.adjust(pvalues, method = "holm")
We can compare the unadjusted with the adjusted values as follows. The changed column indicates which p-values are no longer significant (at 0.05 level) after adjustment.
changed <- ifelse(pvalues < 0.05 & adj.pvalues > 0.05, "X", "")
d <- data.frame(pvalues, adj.pvalues, changed)
row.names(d) <- labels
d
## pvalues adj.pvalues changed
## Competency/Knowledge base (Positively Mentioned) 0.00032 0.00448
## Thoroughness (Positively Mentioned) 0.00023 0.00345
## Thoroughness (Negatively Mentioned) 0.03500 0.30600 X
## Temperament (Negatively Mentioned) 0.02300 0.23000 X
## Cost-consciousness (Negatively Mentioned) 0.03400 0.30600 X
## Interactions with staff (Positively Mentioned) 0.00098 0.01176
## Interactions with staff (Negatively Mentioned) 0.00130 0.01430
## Billing and insurance (Positively Mentioned) 0.12700 0.76300
## Parking (Negatively Mentioned) 0.83800 1.00000
## Black 0.47500 1.00000
## Other 0.34300 1.00000
## Knew Name of Provider (Yes, Reference = No) 0.00043 0.00559
## Responded to comments (Yes, Reference = No) 0.10900 0.76300
## Practice Type (Academic, Reference = Non-academic) 0.12900 0.76300
## Constant 0.16500 0.76300
Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 65–70. https://www.jstor.org/stable/4615733.