Performance of Bias-Corrected Bootstrap Confidence Interval for Parameter k Method in Actor-Partner Interdependence Model

• In Event: Marriott Roundtable Session Five
  In Roundtable Session: Advanced Research Methodology (Table 1)

Abstract

In APIM, four patterns can be examined using parameter k, which is the ratio of partner effect to actor effect (p/a). It can be tested by including a phantom variable in a model and inspecting whether the bootstrapping confidence interval for the k parameter includes 1, 0, or -1. This study aims to examine the performance of the bias-corrected (BC) bootstrap CI for the parameter k method under the various conditions of sample sizes and k ratios. Results revealed that the CI of k was more stable as the sample size increased. Even when the true k ratios somewhat deviated from the values of specific patterns (-1, 0, and 1), more than 80% of the CIs included those, respectively.

Authors

• Junyeong Yang, The Ohio State University
  Presenting Author

• Jiwon Kim, University of Texas at Austin
  Non-Presenting Author

• Minjung Kim, The Ohio State University
  Non-Presenting Author