Errata for Statistics for Linguistics with R, 2nd ed. ===================================================== p. 1, bullet point 3: data, summarize -> data, how to summarize (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 1, last line: they can handle) -> they can handle (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 2, paragraph starting with "Chapter 3": i.e. procedures, in which several potential -> i.e. procedures in which several potential” (thanks to Daria Bębeniec for pointing this out to me) p. 2, lower third: things, Ican only deal -> things, I can only deal (thanks to Laurence Anthony for pointing this out to me) p. 10, Table 3: CLAUSALLY MODIFIED -> CLAUSALLY-MODIFIED (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 11, mid: reduced the minimum age, that the number of -> reduced the minimum age, the number of (thanks to anonymous for pointing this out to me) p. 12, above Figure 3 correlated with independent dependent variables -> correlated with independent and dependent variables (thanks to Laurence Anthony for pointing this out to me) p. 13, above H_0 type 1: both a dependent and and independent variable -> both a dependent and an independent variable (thanks to anonymous for pointing this out to me) p. 24, Table 6 ONJ -> OBJ p. 28, bullet point 2, line 2 from bottom: etc. (Outliers -> etc.) (Outliers (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 28, last par.: , that other probability -> , the other probability (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 33, last full par, line 2 from bottom: where here there -> where there (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 34, caption of Figure 4: All probabilities of possible results -> All possible results (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 36, caption of Figure 6: All probabilities of possible results -> All possible results (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 41, first bullet point the standard normal distribution with z-scores (norm); -> the standard normal distribution (norm) with z-scores; (thanks to Laurence Anthony for pointing this out to me) p. 46, mid of last par.: Since there are two independent variables for each of the two levels -> Since there are two levels for each of the two independent variables (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 49, last par., line 1: do not follow 0 -> do not follow (5) (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 50, line 1: items of Table 13 -> items in Table 13 (thanks to Laurence Anthony for pointing this out to me) p. 53, line 2: in front of -> in front of (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 54, line 11: website (see -> website; see (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 54, line 7 from bottom: Now that every subjects -> Now that every subject (thanks to Laurence Anthony for pointing this out to me) p. 55, line 3: (without double quotes, of course) -> (without double quotes, of course)) (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 55, mid: and blue arrives indicate -> and blue arrows indicate (thanks to Laurence Anthony for pointing this out to me) p. 55, above bullet points: you enter them into and -> you enter them into and (thanks to Laurence Anthony for pointing this out to me) p. 57, line 1 after 6.: files with example files -> files with examples and data sets (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 57, mid: (for _s_tatistics _f_or _l_inguists _w_ith _R_) -> (for _s_tatistics _f_or _l_inguistics _w_ith _R_) (thanks to Laurence Anthony for pointing this out to me) p. 58, line 1: the console with -> the console by (thanks to Laurence Anthony for pointing this out to me) p. 58, line 7: high-lighting -> highlighting (thanks to Laurence Anthony for pointing this out to me) p. 66, last line of code: > numbers -> > numbers1.and.numbers2 (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 69, line 1-2: let R figure out it for -> let R figure it out for (thanks to Poppy Siahaan for pointing this out to me) p. 71, below Figure 12: You get -> You get (with <_inputfiles/02-3-2_vector2.txt>) (thanks to Laurence Anthony for pointing this out to me) p. 71, mid: Now, how do you save vectors into files. -> Now, how do you save vectors into files? (thanks to Peter Hancox for pointing this out to me) p. 74, the bottom second code box > which(x<=7) which elements of x are <= 7?¶ -> > which(x<=7) # which elements of x are <= 7?¶ (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 74, the bottom second code box > which(x>8 | x<3) which elements of x are >8 or <3?¶ -> > which(x>8 | x<3) # which elements of x are >8 or <3?¶ (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 75, last par.: The output of %in% is a logical variable which says for each element of the vector before %in% whether it occurs in the vector after %in%. -> The output of %in% is a logical variable which, for each element of the vector before %in%, says whether or not it occurs in the vector after %in%. (thanks to Peter Hancox for this suggestion) p. 79, beginning of 4: we read in data frames -> we read in data frames (to be discussed below) (thanks to Laurence Anthony for pointing this out to me) p. 79, 2nd to last line of text: mostly the vector -> usually the vector (thanks to Peter Hancox for this suggestion) p. 80, in the middle - 0.992 < interval/level 1 ≤ 3.66; - ? 3.66 < interval/level 1 ≤ 6.34; - ? 6.34 < interval/level 1 ≤ 9.01. -> - 0.992 < interval/level 1 ≤ 3.66; - 3.66 < interval/level 2 ≤ 6.34; - 6.34 < interval/level 3 ≤ 9.01. (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 82, line 2: chang the name -> change the name (thanks to Laurence Anthony for pointing this out to me) p. 86, beginning of 5.2: save it as a comma-separated text file -> tab-delimited text file with the extension .csv (thanks to Laurence Anthony for pointing this out to me) p. 86, last full par.: ; then you choose tabs as field delimiter -> ); then you choose tabs as field delimiters (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 86, bullet point 1: here, too; -> here, too); (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 87, par. before last block of code: with read.delim:, which -> with read.delim, which (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 90, line 2 of second grey box of code Class -> CLASS (thanks to Laurence Anthony for pointing this out to me) p. 90, last line: easy with vector -> easy with vectors (thanks to Peter Hancox for pointing this out to me) p. 91, line 1 after 1st code block b<-a[a$Class=="open",]; b -> b<-a[a$CLASS=="open",]; b (thanks to Peter Hancox for pointing this out to me) p. 91, last two lines: in a spreadsheet software -> in a spreadsheet software program (thanks to Laurence Anthony for pointing this out to me) > p. 92, right above the THINK BREAK you wanted to search the data frame -> you wanted to sort the data frame (thanks to Peter Uhrig for pointing this out to me) > p. 92, right above the THINK BREAK within Class according to -> within CLASS according to (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 93, recommendations to read in tab-delimited files -> to read in comma-separated files (thanks to Laurence Anthony for pointing this out to me) p. 94, first block of code: what to do if this logical expression evaluates to FALSE -> what to do if this logical expression evaluates to TRUE p. 95, last block of code: as often often as -> as often as (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 100, block of code: > peek< function (x, n=6) {¶ -> > peek<-function (x, n=6) {¶ (thanks to Joanna Zaleska for pointing this out to me) p. 118, below third grey box of code: 10.000 -> 10,000 (thanks to Steven Coates for pointing this out to me) p. 118, below third grey box of code: 5.000 -> 5,000 (thanks to Steven Coates for pointing this out to me) p. 136, near the bottom the levels of the first vector in the rows -> the unique values/levels of the first vector/factor in the rows (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 139, code below Figure 31 fil.table -> fill.table (two times, thanks to Rik Vosters for pointing this out to me) p. 153, par. 2: colleages -> colleagues (thanks to Daria Bębeniec for pointing this out to me) p. 162, line 5 of 4.1.1.1: you must some know test like this -> you must know some test like this (thanks to Simone Ueberwasser for pointing this out to me) p. 163, grey block of code: xlab="Tense-Apect correlation" -> xlab="Tense-Aspect correlation" (thanks to Simone Ueberwasser for pointing this out to me) p. 170, mid of last full par.: 0.00000125825 -> 0.000001125825 (thanks to Susanne Flach for pointing this out to me) p. 173, second bullet point: interval/ratio-scaled: the -> interval/ratio-scaled variable: the (thanks to Daria Bębeniec for pointing this out to me) p. 180, last par.: are those represented in Table 24. -> are those represented in Table 25. p. 186, block of code: sqrt(test.Peters$statistic/ sum(Peters.2001)*(min(dim(Peters.2001))-1))¶ -> sqrt(test.Peters$statistic/ (sum(Peters.2001)*(min(dim(Peters.2001))-1)))¶ (thanks to Alvin Chen for pointing this out to me) p. 206, H1: average ofHEDGES -> average of HEDGES p. 211, first grey block of code: Dices <- read.delim(file.choose() -> Dices <- read.delim(file.choose()) (thanks to Simone Ueberwasser for pointing this out to me) p. 217, last grey block of code: ylim=(c(0, 1000)) -> ylim=c(0, 1000) p. 218-221: Given the the test for homogeneity of variances returned a ns result, it would have been didactically more consistent to apply t.test with the additional argument var.equal=TRUE. If the t-test is computed like that, the p-value changes from 0.01619 (with var.equal=FALSE) to 0.01611 (with var.equal=TRUE); the formula for the df changes to n_1+n_2-2. (thanks to Yuliya Morozova for pointing this out to me) p. 221, recommendations box: of this F-test, which -> of this t-test, which p. 229, second line from bottom it may safer to compute -> it may be safer to compute (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 248, line 12: a small a pilot -> a small pilot (thanks to Alvin Chen for pointing this out to me) p. 256, par. 2, line 2 advanatage -> advantage (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 264, first line of first code block: > RTs$FAMILIARITY< factor -> > RTs$FAMILIARITY<-factor (thanks to Alonso Vásquez Aguilar for pointing this out to me) p. 264, the first line of the paragraph right above the last code example ecdf plots). -> ecdf plots), (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 273, mid of last paragraph trials is 0.95^2=0.857375. -> trials is 0.95^3=0.857375. (thanks to Earl Brown for pointing this out to me) p. 276, section heading of 5.2.4 2.4. A linear model with a two categorical predictors -> 2.4. A linear model with two categorical predictors p. 278, the second item in the list at the top (averaging across IMAGEABILITY)’ -> (averaging across IMAGEABILITY); (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 286, mid Part of the reason why the overall model is significant -> Part of the reason why the overall model is insignificant p. 295, (63a): 1/(1+e^x) -> 1/(1+e^-x) (thanks to András Bárány for pointing this out to me) p. 295, Procedure box fewer than 95% of the model'2 absolute -> fewer than 5% of the model's absolute (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 304, mid weak correlation betwween the type -> weak correlation betwwen the type (thanks to Daria Bębeniec and Nathaniel Sims for pointing this out to me) p. 304, last line before 3.2 prefer to preceed main clauses -> prefer to precede main clauses (thanks to Daria Bębeniec for pointing this out to me) p. 304, bottom Since FAMILIARITY has more than 1 df, you use drop1 (or other functions, see the code file) to get one p-value, and FAMILIARITY is highly significant. -> Since CONJ has more than 1 df, you use drop1 (or other functions, see the code file) to get one p-value, and CONJ is highly significant. (thanks to Nina Julich and Daria Bębeniec for pointing this out to me) p. 305, last line of paragraph 1 level of FAMILIARITY -> level of CONJ (thanks to Daria Bębeniec for pointing this out to me) p. 306, par. 3: Fox and Weisberg (2011:239) -> Fox and Weisberg (2011: 239)” (thanks to Daria Bębeniec for pointing this out to me) p. 315, par. 3: (but see Fox and Weisberg (2011: Ch. 6)) -> (but see Fox and Weisberg 2011: Ch. 6) (thanks to Daria Bębeniec for pointing this out to me) p. 319, line 2 of the paragraph below the last code box (and the argument ".*\\." means ‘characters up to and including a period’) -> (and the argument "^.*?\\." means ‘characters up to and including a period’) (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 320, last par., mid: more problematic, though so -> more problematic, though, so (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 322, last full par., line 1: get again get -> again get (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 323, line 2 before last code block: take take -> take (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 328, par. below bullet points: popluation -> population p. 328, last par.: her,e -> here, p. 330: Errror: TEXT -> Error: TEXT p. 333, par. 1 of Section 5.4: use as the above may have made you expect. -> use as you may have expected. (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 333, line 3 from bottom: system, they allow the -> system, allowing the (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 335, par. 2, line 3: chapter. Well -> chapter? Well (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 335, bullet point 2, penultimate line: do no converge -> do not converge (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 335, bullet point 2, mid: with a maximal random-effect structure -> with a maximal random-effects structure (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 336, middle of the page (above the Recommendation(s) box): that make tackling some of these questions more easily -> that make tackling some of these questions easier (thanks to Daria Bębeniec for pointing this out to me) p. 339, Procedure box: (Post-hoc -> Post-hoc (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 339, bottom: clusters with bar (and -> clusters (and (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 342, bulleted list (three times): for aa and bb: -> for aa and bb (thanks to Jae-Woong Choe and colleagues for pointing this out to me) p. 343, middle paragraph: all other measures for ratio-scaled are: -> all other measures for ratio-scaled variables are (thanks to Daria Bębeniec for pointing this out to me) p. 348, line 5 below Figure 81: and the latter very little substructure -> and the latter has very little substructure (thanks to Daria Bębeniec for pointing this out to me) p. 348, line 5 from bottom: The function cluster.stats from the library fpc offers -> The function cluster.stats from the library fpc offers p. 349, line 5: sould -> should (thanks to Daria Bębeniec for pointing this out to me) p. 355: remove the two Divjak & Gries references. (thanks to Daria Bębeniec for pointing this out to me) ######################### last updated 14 July 2019 STG