Using the tab_itemscale() function of the sjPlot package with a discrete response options (A, B, C, D)

Question

I'm analyzing an assessment that has 8 sections, 4 versions of each section that is randomly distributed to respondents.

I'm trying to get at item-analysis using the tab_itemscale() function of the sjPlot package, however it requires me to recode the string variables to numeric.

What is the best practice around this since I don't know how corresponding A, B, C and D responses to a numeric value may affect the understandings of the item.

I don't want to recode to you either getting a 1 or 0 if you got the answer correct because I'd like to understand each individual response and item difficulty.

Thanks for the help!

I created this function:

factorise <- function(x) {
  case_when(x %in% c("A") ~ 1,
            x %in% c("B") ~ 2,
            x %in% c("C") ~ 3,
            x %in% c("D") ~ 4)
}

and applied it the the items of 1 section of 1 version:

WXM_M_v1[,c("M1V1", "M2V1", "M3V1", "M4V1", "M5V1", "M6V1", "M7V1", "M8V1")] <- lapply(WXM_M_v1[,c("M1V1", "M2V1", "M3V1", "M4V1", "M5V1", "M6V1", "M7V1", "M8V1")], factorise)

and then applying the function

tab_itemscale(WXM_M_v1)