Merge pull request #58 from MicheleNuijten/bug-jepg-article

bug fix: don't extract multiple comparisons

R/extract_stats.R

@@ -76,14 +76,29 @@ extract_stats <- function(txt, stat){
 
     test <- extract_test_stats(raw = nhst_raw[i])
 
-    test_stats <- rbind(test_stats, test)
+    # ignore any cases with more than 1 test comparison (e.g., z >= 2)
+    if(nrow(test) > 1){
+      test <- data.frame(test_comp = NA,
+                         test_value = NA,
+                         test_dec = NA)
+    } 
+
+    test_stats <- rbind(test_stats, test)  
 
     # extract p-comparison and p-value
 
     p <- extract_p_value(raw = nhst_raw[i])
 
+    # ignore any cases with more than 1 test comparison (e.g., p >= .01)
+    if(nrow(p) > 1){
+      p <- data.frame(p_comp = NA,
+                      p_value = NA,
+                      p_dec = NA)
+    }
+
     pvals <- rbind(pvals, p)
 
+
   }
 
   # create final data frame ------------------------------------------------------
@@ -122,6 +137,12 @@ extract_stats <- function(txt, stat){
     # and would otherwise break down
     nhst_parsed <- nhst_parsed[!is.na(nhst_parsed$Value), ]
 
+    # remove missing test comparisons or p comparisons 
+    # reason: these are marked as NA in cases where multiple comparisons were
+    # reported. E.g., t(23) >= ..., p >= ...
+    nhst_parsed <- nhst_parsed[!is.na(nhst_parsed$Test.Comparison) &
+                                 !is.na(nhst_parsed$Reported.Comparison), ]
+
     # only return selected stats
     # to that end, rename test-types to match argument stat
     types <- as.vector(nhst_parsed$Statistic)

tests/testthat/test-extract-t-tests.R

@@ -114,3 +114,10 @@ test_that("t-values with a weird minus sign and a space do not result in errors"
     expect_output(statcheck(txt1, messages = FALSE), "did not find any results")
 })
 
+# multiple comparison signs 
+test_that("t-values with multiple comparison signs are not retrieved", {
+  txt1 <- "t(38) >= 2.25, p = .03"
+  txt2 <- "t(38) = 2.25, p >= .03"
+
+  expect_output(statcheck(c(txt1, txt2), messages = FALSE), "did not find any results")
+})