本研究包含2个正式实验，采用眼动加工指标及个体填写的潜在购买意向主观指标，借助眼动仪设备对不同认知风格的个体浏览名人广告及态度产生的整个过程进行记录与分析，为个体眼动加工指标与主观购买意向之间的预测关系寻找实证依据。

实验一探究了不同场认知风格被试在高低两种卷入度条件下其眼动加工特征和潜在购买意向受序列效应的影响，验证了不同认知风格用户在眼动加工特征上均吻合劝说的双加工模型及ELM模型。实验二在实验一的基础之上，将产品卷入度控制在低卷入条件，进一步探讨了代言人与产品匹配程度及展示序列对不同场认知风格被试的潜在购买意向影响，并在低卷入条件下从个体眼动加工特征方面验证了名人产品匹配假设。

通过对两个实验的数据进行分析，可以得出以下结论：

1、不同认知风格的个体其广告内容的眼动加工积极程度和购买意向均受到卷入度因素影响，具体表现为：高卷入条件下个体浏览代言人所展示产品的注视时长和注视点均高于低卷入条件下，低卷入条件下个体的购买意向高于高卷入条件。

2、不同认知风格个体浏览广告内容时，其眼动加工积极程度和购买意向受展示序列因素影响不同。对于场独立型个体而言展示序列并未影响其眼动加工和购买意向，对场依存型个体而言，先向其呈现善于加工的信息线索可促进加工深度和购买意向。

3、代言人与产品匹配与否直接影响了个体对广告页面内容的眼动加工积极程度，但不影响其潜在购买意向。代言人产品匹配时个体眼动加工比不匹配时更为积极。

4、低卷入条件下，不同认知风格个体加工广告内容积极程度及购买意向受展示序列因素调节。眼动加工积极与否受到某一认知风格个体善于加工信息呈现的首因效应影响，个体潜在购买意向受认知风格个体善于加工信息呈现的近因效应影响。

This study consists of two formal experiments, using indicator of eye movement processing and subjective indicator of potential purchase intention by questionnaire, and applying eye movement device to record the whole process of individuals with different cognitive styles browsing celebrity advertisements and attitude generation, so as to find empirical basis for the prediction of the relationship between eye movement processing indicator and subjective purchase intention.                

Experiment 1 explored the influence of sequence effect on eye movement processing characteristics and potential purchase intention of subjects with different cognitive styles under the condition of high and low involvement, and verified that users with different cognitive styles were consistent with the persuasive dual processing model and ELM model in eye movement processing characteristics. In Experiment 2, based on Experiment 1, the product involvement degree was controlled under the condition of low involvement, and the influence of the matching degree of spokesperson and product and the display sequence on the potential purchase intention of subjects with different field cognitive styles was further discussed. In addition, the hypothesis of celebrity product matching was verified from the aspect of individual eye movement processing characteristics under the condition of low involvement.             

The following conclusions are drawn from the analysis of two experimental data:            

1. The active degree of eye movement processing and purchase intention of advertising content of individuals with different cognitive styles are affected by involvement degree factors, which are as follows: under the condition of high involvement, the fixation time and fixation point of products displayed by individual browsing spokesmen are higher than that under the condition of low involvement, and the purchase intention of individuals under the condition of low involvement is higher than that under the condition of high involvement.           

2. When individuals with different cognitive styles browse advertising content, their active degree of eye movement processing and purchase intention are affected by display sequence factors. For the field independent individuals, the display sequence did not affect their eye movement processing and purchase intention. For the field dependent individuals, first presenting processing information clues to them can promote processing depth and purchase intention.          

3. Whether the spokesperson matches the product directly affects the individual's active degree of eye movement processing of the content of the advertising page, but does not affect their potential purchase intention. The individual eye movement processing is more active when the spokesperson is matched with the product than when the spokesperson is mismatched.

4. Under the condition of low involvement, the positive degree of processing advertising content and purchase intention of individuals with different cognitive styles are regulated by display sequence factors. Whether eye movement processing is active or not is affected by a cognitive style individual being good at processing information presentation under the first cause effect, and individual potential purchase intention is affected by cognitive style individual being good at processing information presentation under the proximate cause effect.