¹ The Democrats held 26 seats and the Republicans 24. The Democratic majority was the first in the Indiana Senate since 1939.

² χ ² = 15.7, df = 1, p <.001.

³ χ ² = 16.34, df = 1, p <.001.

⁴ The interaction scale is an adaptation from a scale presented by Caplow, Theodore and Forman, Robert, “Neighborhood Interaction in a Homogeneous Community,” American Sociological Review, Vol. 40 (June, 1950), p. 358 Google Scholar. An elaborate description of its reliability and usefulness can be found in a work by Theodore Caplow, Sheldon Stryker, and Samuel E. Wallace, The Urban Microcosm: Neighborhood Structure in a Changing Metropolis, manuscript in preparation. Establishing the reliability of the neighborhood interaction scale, of course, only indirectly supports the use of a similar scale in a legislative situation. In the adaptation of the scale it was feared that the seasonal nature of legislative activity would raise the interaction ratings during the latter part of the interviewing period. Consequently, three weeks were allowed to elapse before beginning the interviews, in the hope that the social patterns would stabilize by that time. The evidence indicates that this was the case. The first ten interviews were compared with the last ten and no rise ocmirred in the average scale value for the ratings made by the latter ten. By necessity, no interviews were conducted during the last week of the session; however, for each group of ten interviews the party distribution and experienced-inexperienced distribution reasonably approximated the distributions in the entire Senate and for all those interviewed.

⁵ χ ² = 17.8, df = 1, p <.001.

⁶ Ties occurring at the lowest boundary of categories were broken randomly. For more precise purposes the index would not require equal groups of ten or any other number, but for the sake of clarity it is helpful to think of a .7 index value as an overlap of seven senators out of ten for two separate sets of data.

⁷ Hovland, Carl I., Janis, Irving L., and Kelley, Harold H., Communication and Persuasion (New Haven, 1953)Google Scholar. The authors emphasize the importance of personality in determining influence patterns.

⁸ This classification is based upon an accumulative quartile ranking. Let Q1 be the highest ranking and Q4 the lowest for both bill success and formal position success. A legislator's accumulative ranking can then range from the sum of 2 to the sum of 8. The line between high and low was drawn at that class boundary which most closely approaches the median (between the sums of 4 and 5).

⁹ rϕ = √/N The phi coefficient has more meaning when it is considered in conjunction with chi-square (χ ²). When N is constant rϕ varies directly with χ ²; therefore, certain values of rϕ reflect significant values of <.05, rϕ>.277; when p<.02, rϕ>.329; when p<.01, rϕ>.365; when p<.001, rϕ>.466.

¹⁰ In the Indiana Senate no bill may have more than two named sponsors. The sponsors generally explain their own bills on the floor and quite often in committee.

¹¹ χ ² = 2.00, df = 1, p>.05.