Trait and State Approaches to Explaining Argument Structures

Abstract

This study examines how an individual's argumentative communication traits and involvement, a situational factor, account for the way the individual structures his or her written argument. Research questions were posited to ask which of the 3 models—the Interaction Model, the Mediation Model, and the Simple Main Effects Model—best predicts the use of macro- and micro-structures of written argument. Logistic and multiple regression analyses were conducted to analyze 229 usable responses collected from Japanese college students. The results revealed that the Mediation Model and the Simple Main Effects Model are viable models. Argumentative traits and involvement interrelate to explain argument structures, revealing a somewhat complex but interpretable mechanism.

Keywords:

• Argumentativeness
• Argument Structures
• Involvement

The research on which this essay was based was funded by the Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research (C) 2052048408. I thank Hisae Hashimoto, Kazuko Inoue, Masatoshi Miyashita, Miori Nagashima, and Eiko Tsuchida for data collection; and Azusa Sato for her help in coding the transcripts.

Notes

Note. N = 229, except for climactic versus anticlimactic macro-structure for which N = 216. ARGap = approach argument; ARGav = avoid argument.

^a Motivation to ARGap was measured by 10 items on a 5-point scale, ranging from 1 (the description hardly ever applies) to 5 (the description almost always applies). The same was true for ARGav.

^b Involvement was measured by three items on 5-point scales, ranging from 1 (almost never) to 7 (very often), 1 (not at all) to 7 (to a very high degree), and 1 (not at all important) to 7 (very important).

^c This is a dichotomous variable; horizontal macro-structure was coded as 0, whereas vertical macro-structure was coded as 1. Of the 229 responses, 43 had horizontal macro-structure in which the nuclear claim (NC) is not supported by any reason or vertically subordinate (VS) statement; 186 responses had vertical or linear macro-structure, which contains support or reasoning or VS statements for the NC. The 43 arguments with horizontal macro-structure included 13 arguments, which consisted of only one unit coded as NC.

^d This signifies the relative location of the NC, represented by the order in which the NC appears in the argument, divided by the total number of units. Thirteen one-unit arguments were excluded from the analysis because it was hard to decide on the climacticity of arguments consisting of only one unit.

^e This is a categorical variable. When an argument had no serial-type micro-reasoning structure, it was coded as 0. When an argument had one or more serial-type micro-reasoning structure, it was coded as 1.

^f The proportion of the compound-type micro-reasoning structures in an argument-to-argument length.

Note. N = 229, except for climactic versus anticlimactic macro-structure for which N = 216. ARGap = approach argument; ARGav = avoid argument.

*p < .05. **p < .01 (two-tailed).

Note. N = 229. The Sobel test indicated that the drop in the direct effect of ARGap on the dependent variable was significant (z = 2.85, p = .002). ARGap = approach arguments; ARGav = avoid arguments.

^a When involvement was added to the equation in Step 2, the effect of ARGap on the dependent variable diminished.

*p < .05. **p < .01 (two-tailed).

Note. N = 216. ARGap = approach arguments; ARGav = avoid arguments.

*p < .01 (two-tailed).

Note. N = 229. ARGap = approach arguments; ARGav = avoid arguments.

*p < .01 (two-tailed).

Note. N = 229. ARGap = approach arguments; ARGav = avoid arguments.

*p < .01 (two-tailed).

Note. Argument A has (a) a vertical macro-structure because its nuclear claim (NC) is supported by one or more reasons. Its (b) climacticity score is .17 = 1 (the location of the NC)/6 (the total number of units). The argument's (c) use of the serial-type micro-reasoning structures is 1, and its (d) proportion of the compound-type micro-reasoning structures to argument length is 0 = 0/6. Argument B has (a) a vertical macro-structure because its NC is supported by one or more reasons. Its (b) climacticity score is 1.0 = 6 (the location of the NC)/6 (the total number of units). The argument's (c) use of the serial-type micro-reasoning structures is 0, and its (d) proportion of the compound-type micro-reasoning structures to argument length is .33 = 2/6. Argument C has (a) a horizontal macro-structure because its NC is not supported by any reason. Its (b) climacticity score is 1.0 = 5 (the location of the NC)/5 (the total number of units). The argument's (c) use of the serial-type micro-reasoning structures is 0, and its (d) proportion of the compound-type micro-reasoning structures to argument length is 0 = 0/5. HC = horizontally continuing; -QU = qualification; VS = vertically subordinate; -QL = quasi-logic; CE = cause and effect; -CL = clarification; CI = circumstance.

Power analysis was conducted using PASS 2008 (Hintze, 2008), a power analysis and sample size software program, for the logistic regression analysis that was used to test the categorical dependent variables. For a sample size of 250 observations, with the level of significance set at .05, with a moderate odds ratio, which is an index of effect size in logistic regression, being 2.0, the power of the analysis was .99. Odds ratio of 2.0 in this case meant a change in the probability of Y, the binary dependent variable, equals 1 from the value of 0.8 (P1) at the mean of X, the independent variable, to 0.889 (P2) when X is increased to 1 SD above the mean—that is, it is [P1/(1 – P1)]/[P0/(1 – P0)]. Power analysis was also conducted for the multiple regression analyses that were used to test for the rest of the dependent variables. For a sample size of 250 observations, with the level of significance set at .05, and with a moderate effect size of f ² = .15, the power of the analysis was above .99.

Additional information

Notes on contributors

Shinobu Suzuki

Shinobu Suzuki (Ph.D., University of Minnesota, 1994) is an associate professor in the Graduate School of International Media, Communication and Tourism Studies at Hokkaido University.