A major release that tailors the package to the 4th edition of Maxwell, Delaney, and Kelley, Designing Experiments and Analyzing Data: A Model Comparison Perspective (2027, Routledge).
chapter_12_exercise_22 (C12E22)
and chapter_12_exercise_23 (C12E23), the Hu,
Antony, Creery, Vargas, Bodenhausen, and Paller (2015)
sleep-and-implicit-bias study (Exercise 22 is the 2 by 2 cued/uncued by
prenap/postnap design for 40 participants; Exercise 23 adds the one-week
delayed measurement for the 38 participants who returned); and
chapter_12_exercise_25 (C12E25), the Kroes,
Tendolkar, van Wingen, van Waarde, Strange, and Fernandez (2014)
electroconvulsive-therapy memory study (three groups, reactivated and
non-reactivated story scores; its variables are cond,
reactivated, and nonreactivated, matching the
names the exercise gives). The same Kroes study supplies the one-way
data of chapter_4_exercise_21. Each reproduces the results
reported for its exercise.chapter_12_exercise_24 (C12E24), the
Carnes, Lickel, and Janoff-Bulman
chapter_12_exercise_26
(C12E26), the Gibson, Radvansky, Johnson, and McNerney
(2012) grapheme-color synesthesia study (48 nonsynesthetes and 10
synesthetes, with input-output correspondence and recall accuracy for
high- and low-frequency words); and chapter_12_exercise_27
(C12E27), the Bray and Maxwell (1995) reports of negative
life events from the mother, father, and child of 95 nuclear families
and 98 stepfamilies. Each reproduces the results reported for its
exercise.chapter_14_table_13 (alias
C14T13): the longitudinal randomized design introduced in
Chapter 14 (100 participants, 44 treatment and 56 control, measured at
baseline and two post-intervention occasions). These data underlie
Tables 14.13 and 14.14 and are analyzed again with mixed-effects models
in Chapter 15 (Tables 15.15 and 15.16). Stored in the long format
expected by the Chapter 15 code.chapter_1_table_1) and its short alias (e.g.,
C1T1), and the numeric coding of variables is preserved so
the book’s model-comparison and contrast examples reproduce
exactly.chapter_7_table_15 becomes chapter_7_table_16,
chapter_7_table_23 becomes chapter_7_table_24,
chapter_9_table_11 becomes chapter_9_table_12,
chapter_12_table_21 becomes
chapter_12_table_29, chapter_13_table_5
becomes chapter_13_table_6,
chapter_13_table_12 becomes
chapter_13_table_10, and the split-plot reaction-time
tables chapter_14_table_8 and
chapter_14_table_10 become chapter_14_table_7
and chapter_14_table_9. Only the names changed; the data
values are identical. The short aliases were renamed to match (for
example, C7T15 becomes C7T16).chapter_14_table_8 (the
M and D scores for the data in Table 14.7) and
chapter_14_table_10 (the M, D1, and
D2 scores for the data in Table 14.9), each also under its
short alias (C14T8, C14T10).chapter_13_table_14 (and its alias
C13T14). Its slope and linear-trend scores correspond to a
3rd-edition table that is not part of the 4th edition; the linear-trend
values it tabulated remain available as the Linear column
of chapter_13_table_10.chapter_4_table_7 (and its alias
C4T7). It held data for a 3rd-edition demonstration that
the sums of squares of orthogonal contrasts are additive while those of
nonorthogonal contrasts are not; in the 4th edition that demonstration
moved to the online supplement (designingexperiments.com/supplements),
so the data set no longer corresponds to a printed table. The 4th
edition’s printed Table 4.7 is the Exercise 11 data, which is already
shipped as chapter_4_exercise_11.chapter_7_table_9 (and its
alias C7T9). It had held a 36-row combined data set (the
Table 7.5 blood-pressure data with a sixth observation appended to each
cell). The 4th edition’s Table 7.9 is instead the six
additional observations themselves (one per cell), so the data
set now holds those six rows. Stacking them onto
chapter_7_table_5 with rbind() recreates the
six-per-cell data whose cell and marginal means are reported in Table
7.10.chapter_4_exercise_18 (C4E18, four-therapy
agoraphobia fear scores), chapter_4_exercise_21
(C4E21, the Kroes et al. (2014) three-group ECT memory
data), chapter_6_exercise_11 (C6E11,
drug-dosage maze errors), chapter_7_exercise_21
(C7E21, the oxytocin-by-gender trust study), and
chapter_13_table_5 (C13T5).
chapter_13_table_5 is Table 13.5 of the 4th edition (“Two
Sets of Hypothetical Data to Illustrate How a Determinant Reflects
Generalized Variance”); it holds only the two D (difference) variables
for each of the two data sets, because the book gives only those derived
scores and not the raw data they were computed from. (In the 4th edition
the 3rd-edition Table 13.5 was renumbered to Table 13.6, shipped as
chapter_13_table_6.)chapter_7_exercise_18 (and
its alias C7E18). Every value had been 40 (thousand) too
low relative to the 4th edition’s Exercise 18 table; the salaries now
match the book exactly (Assistant-rank females 73, 76, 75, 78, 82, 77;
and so on). The level and gender codes are
unchanged.chapter_14_exercise_10
(and its alias C14E10). The data are the Kosslyn
grids-versus-brackets by left-versus-right hemisphere reaction times
(the same data as chapter_12_exercise_9), but the columns
had been mislabeled Angle0, Angle4,
Angle8, Group (copied from the split-plot
angle example); they are now GridLeft,
GridRight, BraceLeft, BraceRight.
The values are unchanged.chapter_15_exercise_19 (and its alias C15E19)
from Cognitive73 to Cognitive72. The three
columns record ability at ages 70, 72, and 74 (as the companion
chapter_15_exercise_19_univariate records with
Age = 70/72/74), so the middle column’s name was corrected
to match; the data values are unchanged.tutorial_1_table_1 (and its alias
T1T1). A header-less import had turned the first score (22)
into the column name (X22) and dropped that observation,
leaving 102 rows. The data set now holds the full 103 Beck Depression
Inventory scores in a column named BDI, reproducing the
statistics reviewed in Tutorial 1 (sum 1,622, mean 15.748, median 13,
mode 9, range 0 to 43, skewness .629, kurtosis -.243).chapter_16_table_1 (and its
alias C16T1) to Trainee, Gender,
Severity, and reordered chapter_16_exercise_5
(C16E5), which analyzes the same Table 16.1 data, to match.
The data values are unchanged.remotes::install_version("AMCP", "1.0.2").Variables lists in several help pages
that had been copied from the wrong data set, so the documented
variables now match the shipped columns: chapter_3_table_3,
chapter_7_table_11, chapter_10_table_9,
chapter_11_table_20, and
chapter_13_table_10.inst/WORDLIST for the package’s domain
vocabulary.chapter_13_table_10 examples now show how its
linear and quadratic D variables are formed from the raw three-level
data of chapter_13_table_2 using the orthogonal-polynomial
coefficients of Appendix Table A.10.chapter_7_table_9 examples now show how its six
additional observations combine with chapter_7_table_5 to
reproduce the cell and marginal means of Table 7.10.Feedback and Drug variables of
chapter_7_table_5 and chapter_7_table_9
informative level labels (Feedback: 1 = biofeedback present, 2 = absent;
Drug: 1 = X, 2 = Y, 3 = Z), matching the labels used in the book’s Table
7.10.Note on its help page
identifying the change (and, for a renumbered or reused name, the
corresponding 3rd-edition name), so readers coming from the 3rd edition
are not misled by a shifted or reused table number.testthat (edition 3) test suite that loads
every data set, checks that the long and short aliases agree, and
reproduces several numeric results reported in the book (the Chapter 3
one-sample t test, the Chapter 13 trend-contrast omnibus
F, and the Chapter 9 Pygmalion heterogeneity-of-regression
slopes).factors-and-coding, with recipes for
attaching factor labels to the numerically coded variables without
changing the canonical data, including a worked example confirming that
relabeling reproduces the book’s results.CITATION file.news(package = "AMCP") for the list of changed tables, and
directing 3rd-edition (2018) users to the archived
remotes::install_version("AMCP", "1.0.2").URL field.