¹ The list of papers and the names of editors were compiled from the 1932 edition of Ayer and Son's American Newspaper Annual and Directory.

² See Vol. 26, no. 3 (June, 1932), pp. 527–544. Directions for marking are important, and may be observed from the heading of the ballot form used, (see bottom of next page.) All fifty-seven statements of the candidate scale were included, and a blank line number 58 was added for writing in additional statements desired.

³ The balloted opinions were those of the individual editors, not necessarily those expressed in the editorial columns of their papers. Political affiliation of the papers was taken as means for illuminating classification of editors and their individual opinions.

⁴ Some were editors of two papers, of different party affiliations.

⁵ Of the 133, a total of 130 checked scaled statements as well as indicated their choice for President. Only 111 checked scaled statements respecting their candidate's strongest opponent in addition to naming him.

⁶ In this section, the numbers in parentheses are the percentages which the given figures are of those for the entire group addressed.

⁷ Calculation of “least circulation” is based on known circulation in the case of identified editors and least circulation of a paper in a city represented by an unidentified editor. Use of largest circulation of a paper in a city represented by an unidentified editor gives the larger figure.

⁸ Congressional Directory, 72d Cong., 1st Sess., January, 1932, pp. 592–606Google Scholar.

⁹ These are being used in a study of the editorial product of selected newspapers during the campaign.

¹⁰ Industrial Belt includes the following groupings mentioned above: New England, Middle Atlantic, and East North Central. East has reference to the combined New England and Middle Atlantic sections; North Central, the combined East North Central and West North Central; and West refers to the combined Mountain and Pacific sections.

¹¹ Detailed analysis of the poll can be secured in mimeographed form from the author.

¹² In this section, the figures in parentheses are the percentages of the stated group of editors who endorsed a statement.

¹³ Thirty-eight per cent of the anti-Hoover editors deemed this statement to be true of the Republican candidate.

¹⁴ The probable errors of the scale value of the statement landmarks of the candidate and of other similar scales are compared in an article on “Measuring the Severity of the Third Degree,” shortly to appear in the Journal of Criminal Law and Criminology.

¹⁵ The average central tendency of the four editors favoring Thomas was 8.0; while that of the two editors favoring Smith was 8.4. The sample expressing these preferences was much too small to justify treatment or further statement.

¹⁶ A mimeographed copy of the more detailed analysis can be secured from the author.

¹⁷ Only three editors submitted identifiable ballots expressing choice for a candidate opposed by the politics of the editors' papers. If any other editors balloted in opposition to the political affiliation of the paper they represented, they submitted unidentifiable ballots summarized in this column.

¹⁸ In this connection, it may be of interest to note that the Pearsonian coefficient of correlation between the proportion of balloters favoring a candidate in the Digest poll and the proportion favoring a candidate in the official vote of November 8 was +.650±.030, when the data was classified by intervals of five per cent. The coëfficient was +.598±.034 when the data was classified by intervals of one percent. These correlations treated only the proportions from cities involved in this poll and for which the official returns could be secured. Correlations reported in the text were based on the finer classifications.