Abstract
In general, people like to hold correct attitude. In order to achieve the goal of subjective correctiveness, people will take effort to process information and they will try to remove the bias as long as the bias is perceived by them. In this study, we discuss the argument sufficiency as a moderator under elaboration likelihood model （Petty & Cacioppo, 1986）and judgmental correction situation. In addition, we exam if people will identify issue-irrelevant source with attractive characteristic as a bias under ELM model. We try to verify the result of source effects under ELM. Under high elaboration conditions, people are more likely to be influenced by argument quality rather than by source attractiveness. In other words, the prediction of ELM is that high involvement people deal with either attractive or average source would lead to no different attitude toward product. But under low elaboration conditions, people may consider source is III a simple positive or negative inference, and their judgments toward the product depending on the cue. However, a study proposed by Kang and Herr in 2006 conflicts with ELM model. They find that people with excessively high cognitive resource will remove the influence generated from issue-irrelevant source. That is, extremely high involvement people’s attitude will be more favorable toward the product when the endorser is an average person rather than the attractive endorser. In their study, they don’t manipulate different levels of high involvement, and then solely make their conclusion about the negative source effect resulting from fairly high involvement. Besides, they also neglect to compare argument quality under different conditions. All participants receive only weak arguments. We are unable to exam how weak these arguments are. These arguments are likely to be strong enough for participants to judge the target. Therefore, in our study, we propose argument sufficiency as a moderator which is likely to lead to negative source effects. Additionally, we base on Kang and Herr’s study to improve our experiment. Thus, based on bias correction model, we manipulate the argument sufficiency as a moderator in existing ELM model to explain the contradiction about source effects. If we provide not only sufficient but also strong arguments to high involvement individuals, their judgment won’t be affected by source attractiveness, which consists IV with ELM. However, if product-irrelevant endorser solely with attractive characteristic is presented without any arguments for high involvement individuals, because they can only make their judgment by source, the correction will trigger. The implication of the expected result means bias correction would be triggered under insufficient situation for high involvement individuals.