Analysis of Methods of Disposition of Representation Cases Closed, Fiscal Year 1956

p. 185. Grid is 9 columns: stub plus four case-type panels (All R, RC, RM, RD), each with a "Number of cases" and a "Percent of cases closed" column. The stub heading "Method and stage of disposition" spans both header levels and is repeated in each. Decimal points: the scan is a coarse bitonal image in which many decimal points in the percent columns failed to print/reproduce, so they appear as a blank gap (e.g. RC "100 0" in the total row, All R "18 2" and RD "7 2" in the first "Before notice of hearing" row, All R "24 3" and RC "22 9" in "Withdrawn", RC "15 7" and RM "12 1" in "Board-ordered election", and similar cells elsewhere). These are unambiguously one-decimal percentages in a column where every other value is printed with a point, so I transcribed them with the decimal point restored rather than as a space. No digits were altered. "By Board decision", All R cases / Number of cases: printed as a raised 2 immediately before 273; read as the footnote-2 marker and recorded as 273^2. Indentation is not encoded in the grid: the indented rows ("Before notice of hearing", "After notice of hearing, before hearing opened", "After hearing opened, before Board decision", "After Board decision and direction of election", "After postelection decision") are subordinate to the preceding flush-left method row (Consent election, Stipulated election, Withdrawn, Dismissed), and several of those labels repeat verbatim under more than one method. Arithmetic as printed: the method-level numbers for All R cases (2,131 + 1,757 + 1,959 + 975 + 1,240 + 8) sum to 8,070, matching the printed total; no misprints detected. Nothing appears cut off by the crop.