BooksDavies · 1956
The Design and Analysis of Industrial Experiments
- Matches
- Very close
- Close
- Differs
Factorial gives the book's answers, with a few differences. Where they differ, this page says why: some are misprints in the book, and for some the reason is not known yet.
Examples
Each example, and how closely Factorial gives its answers.
Pages 1 to 99
Pages 100 to 199
- Sampling and testing an organic chemical: 9 batches, 3 samples a batch, 2 tests a sampleExample · pp. 105-114, 133-134Misprint found
- Bulk density of precipitated chalk: 11 laboratories, 2 chalks, 3 repeat testsExample · pp. 115-121, 136-139Close
- Testing dyestuffs: 6 batches, each tested 6 times by each of 2 laboratoriesExample · pp. 121-124, 140-142Very close
- Sampling and analysis of a fertiliser: 8 batches, 2 samples a batch, each analysed by 2 methodsExample · pp. 124-130, 142-144Matches
- Loss of product in the mother liquor: five blends of acetanilide in three randomised blocksExample · pp. 153-159, 192-193Differs
- Wear testing of rubber-covered fabric: four materials, four runs and four positions in a 4 x 4 Latin squareExample · pp. 163-166, 194-195Matches
- Persistence of insecticide dusts: seven methods of mixing, forms of active ingredient and diluents in a 7 x 7 Latin squareExample · pp. 166-169, 196Very close
- Annealing copper tubes: eight lots and a dummy factor in an 8 x 8 Graeco-Latin square on the annealing jigExample · pp. 171-174, 196-197Not yet explained
- Strength of synthetic yarn: times of spinning, days of drawing and two machine positions in four 3 x 3 Graeco-Latin squaresExample · pp. 176-185, 198Differs
Pages 200 to 299
- Road tests on tyres: four tread compounds, three to a tyre, in a balanced incomplete block designExample · pp. 200-213, 245-247Matches
- Abrasion resistance of rubber: five materials in five runs of the four-position wear tester, a Youden squareExample · pp. 215-218Matches
- A two-way table with five values missing, analysed by estimating them and by the exact methodExample · pp. 236-242Very close
- A 2 x 3 x 4 factorial in duplicate: two methods, three temperatures and four batches of raw materialExample · pp. 248-250Matches
- A replicated 2 x 2 factorial, used to define main effects and interactionExample · pp. 254-255Matches
- Yield of a nitration process: time of addition of nitric acid, time of stirring and heel, a 2^3 factorialExample · pp. 258-269Matches
- Yield of an isatin derivative: acid strength, reaction time, amount of acid and temperature, a 2^4 factorialExample · pp. 274-280Very close
- Wear resistance of vulcanised rubber: five fillers, three pretreatments and four raw rubbers, a 5 x 3 x 4 factorialExample · pp. 291-296Very close
- Corrosion of aluminium alloys: nine alloys exposed at four sites, each plate scored by four observersExample · pp. 296-303, 340-344Matches
Pages 300 to 399
- A 4 x 5 factorial with one qualitative and one quantitative factor, analysed by orthogonal polynomialsExample · pp. 305-312Not yet explained
- Penicillin inoculum: five sugars at four concentrations, the responses adjusted by a controlExample · pp. 312-318Very close
- Yield of a direct cotton dyestuff: concentration, volume of free water and amount of a second reactant, a 3^3 factorial in duplicateExample · pp. 332-335, 352-361, 431-435Not yet explained
- Yield of an organic chemical: quality and amount of ammonium chloride and two units, a 2^3 in duplicate with ABC confounded with four lots of base materialExample · pp. 372-378, 410-413Very close
- Penicillin in surface culture: corn steep liquor, lactose, precursor, sodium nitrate and glucose, a 2^5 in two weeks (and in four blocks of eight)Example · pp. 383-387, 394-395, 416-418Very close
Pages 400 to 499
- Filtration in the preparation of a dyestuff: four factors in eight runs, a 2^(4-1) with D = ABCExample · pp. 454-457, 491Matches
- Yield of a medicinal product: five factors in eight runs, a 2^(5-2) with D = -AC and E = ABExample · pp. 457-459, 492Very close
- Quality of a basic dyestuff: five factors over three stages in sixteen runs, a 2^(5-1) with E = -ABCDExample · pp. 462-465, 493Very close
- Yield of penicillin: five factors in a 2^(5-1) with E = ABCD, in four fermenters confounding BC, BE and CEExample · pp. 466-469, 493-494Matches
Pages 500 to 599
- Yield of a laboratory reaction, first stage of steepest ascent: five factors in eight runs, a 2^(5-2) with x4 = x1x2x3 and x5 = -x2x3Example · pp. 511-514Differs
- Yield of a laboratory reaction, second stage of steepest ascent: a 2^(5-2) with x4 = x1x2x3 and x5 = x1x2, around trial (10)Example · pp. 515-518Matches
- Yield in a near-stationary region: a 3^2 factorial, a second-degree fit and its canonical analysisExample · pp. 518-529Close
- Orthogonal composite designs for two to five factors, one centre runDesign · p. 534Matches
- Yield of C in a reaction A + B -> C + D: a 2^3 factorial completed to a composite design (alpha = 2), with canonical analysisExample · pp. 541-550Close
- A hypothetical 2^2: the yield of a reaction, to define main effects and interactionExample · pp. 584-585Matches
Misprints
Misprints found in the book, by page. Each shows what the book prints and what its own data give.
Anova sum sq: Batch·Samplep. 110
Book 25.32Data 25.30
The book gets 25.32 by subtraction from a total of 838.37 (Table 4-11); its own Appendix 4C gives the total as 838.35, and the sum of squares between samples is then 25.30, as the library finds
Ss totalp. 110
Book 838.37Data 838.36
Table 4-11 prints 838.37; Appendix 4C (p. 133) works the same total out as 838.35, which the library agrees with
Anova sum sq: Np. 334
Book 1723Data 1722
The linear contrast of the N totals, (2,408 - 2,159)^2 / 36 = 1,722.25, as the book's own mean square, 1,722, has it; it prints the sum of squares as 1,723
Not yet
Answers the book prints that Factorial cannot compute yet, by what it lacks.
- The sum of squares between block totals, ignoring treatments; the library reports each term adjusted for the otherp. 208
- The correction for the mean, n times the mean squared (699.1²/9); the library's sums of squares are about the meanp. 521
- The uncorrected total, Σy²; the library's total is about the mean (78.48 = 78.06 + 0.42)p. 521