Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Education and Decision-Making: An Experimental Study on the Framing Effect in China

Education and Decision-Making: An Experimental Study on the Framing Effect in China fpsyg-08-00744 May 9, 2017 Time: 15:44 # 1 ORIGINAL RESEARCH published: 11 May 2017 doi: 10.3389/fpsyg.2017.00744 Education and Decision-Making: An Experimental Study on the Framing Effect in China 1,2 Wen Fan * 1 2 School of Public Administration, Nanjing University of Finance and Economics, Nanjing, China, Center for Social Security Study, Wuhan University, Wuhan, China China’s higher education expansion policy has been in effect for almost two decades. Under this policy, a growing number of youths have gained access to higher education, which aims to train students to be more rational. This study examines human rationality at a Chinese college through an experiment based on the risky-choice framing effect. The basic results show no classical framing effect with regard to individual decisions for the entire sample in a benchmark setting. However, when the participants’ roles were manipulated and subsamples were investigated, a significant framing effect was found that appeared to be role-related and that varied by sex. These results help to elucidate evaluations of the effects of China’s higher education policy and may assist in guiding further policy reforms. Keywords: framing effect, decision making, heterogeneity, higher education policy, China Edited by: Michael S. Dempsey, INTRODUCTION Boston University, USA Reviewed by: People make many decisions and judgments in daily life, the explanations for and predictions of Renata Melinda Heilman, which are often based on the assumption of human rationality. However, as Tversky and Kahneman Babes-Bolyai ¸ University, Romania (1986) stated, “Alternative descriptions of decision problems often lead to different preferences, Azizuddin Khan, contrary to the principle of invariance that underlies the rational theory of choice. Violations of Indian Institute of Technology this theory are traced to the rules that govern the framing of decisions and to the psychophysical Bombay, India principles of evaluation embodied in prospect theory” (p. S251). This decision bias is well known as *Correspondence: a risky-choice framing effect, in which choices involving gains are frequently risk-averse, whereas Wen Fan choices involving losses are frequently evidence of risk-taking. The framing effect is commonly nonpareilian@gmail.com considered to be one of the most severe violations of normative utility axioms and is therefore a strong indicator of irrationality. Thus, individuals’ preferences will reverse or shift (bidirectional Specialty section: This article was submitted to or unidirectional framing effects, respectively) when the same problem is framed in different ways. Educational Psychology, Classic bidirectional framing effects frequently lead to irrational reversals in risk preferences under a section of the journal different framing conditions. However, as Wang (1996b) defined the term, “[the] unidirectional Frontiers in Psychology framing effect involves no preference reversal but a shift to a more extreme risk preference: : : if the Received: 29 August 2016 predominant preference is unidirectionally risk-averse under both framing conditions, it is even Accepted: 24 April 2017 more risk-averse when positively, as opposed to negatively, framed. Similarly, if the predominant Published: 11 May 2017 preference is unidirectionally risk-seeking under both framing conditions, it is even more so under Citation: a negative frame” (p. 5). Fan W (2017) Education This study investigates human rationality through an experiment that examines the risky- and Decision-Making: An choice framing effect. The participants were undergraduate students from two different majors Experimental Study on the Framing at a Chinese college. Many studies have been conducted with US college students (e.g., Tversky Effect in China. Front. Psychol. 8:744. doi: 10.3389/fpsyg.2017.00744 and Kahneman, 1981; Levin et al., 2002). However, recent studies in the Chinese context have been Frontiers in Psychology | www.frontiersin.org 1 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 2 Fan Framing Effect in China limited by small sample sizes or the lack of a detailed analysis. MATERIALS AND METHODS For example, Zhang and Miao (2008) conducted similar lab experiments, but the authors focused on the experiment per Participants and Design se. Zhang et al. (2008) included both military and civilian The participants included 351 students (65% female, female mean students, but the authors did not explore heterogeneity effects. age D 21.6, male mean age D 21.8) studying in two majors Compared to the sample sizes of similar experiments conducted (mean age D 20 for social science and mean age D 22 for in developed countries, this study’s larger sample size (N D 351) engineering) at a 4-year college in eastern China; the students yields more power for testing the reliability of the framing received class credit in exchange for their participation in the effects. In addition, recruiting participants from diverse majors experiment. The responses were anonymous, and the instructions allowed a further exploration of the different individual factors specified that although there was no “correct” answer to the that have been employed as determinants (e.g., self-esteem, problem, careful thinking would be highly appreciated. The numeracy, and biological conditions). Thus, this lab experiment students were not permitted to speak with one another during at a Chinese college contributes new evidence to the literature the experiment. All questions were presented in written form, by allowing identification of the perspective- and sex-dependent and the split-ballot questionnaire was administered in two choice patterns that reveal that human cognitive mechanisms classrooms. The study was approved by the research ethics are sensitive to the internal biological status of the information- committee of Nanjing University of Finance and Economics, and processing organism. conforms to the ethical principles of the Declaration of Helsinki (World Medical Association, 2013). Before administering the In addition, as the Chinese Higher Education (HE) expansion policy has been in effect for almost two decades, the public experiment, one of the instructors translated the principle of is highly interested in the linkage between this nationwide informed consent according to Standard 8.02 of the American education policy and the social and economic changes it has Psychological Association’s (APA) new ethics code into Chinese brought to Chinese life and society. There is substantial literature and explained it to all of the participants. The experiment had on how education in different disciplines relates to people’s a two-way mixed design in which an alpha level of 0.05 was choices (see Chambaere et al., 2013, for end-of-life research; Kidd used for all t-tests. The related measures independent variable et al., 2015, for pro-environment behavior analysis; Nafría et al., was the within-participants manipulated role (general, medical 2015, for an eParticipatory decision-making study; and Petroman worker, and president). The unrelated measures independent et al., 2015, for food consumption). variable was the between-participants decision-making domain Most importantly, there is an increasing body of literature (gain and loss). Following Harris (2008), a pilot test was showing that individual rationality or decision-making power administered on a randomly selected smaller group that allowed appears to be positively related to education level. Anderson the order of roles to be counterbalanced such that their order et al. (2017) found strong evidence that a husband’s allocation of presentation varied among the participants. Trial 1 was of decision-making authority to his wife varies according to his under the positive frame, in which one group of subjects wife’s age and education. The authors stated that “on average, (Group 1) first received the Q2 scenario (imagining themselves intra-household accord over which spouse holds decision- as a medical worker) and a second group (Group 2) first making authority is more likely in households where women received the Q1 scenario (just acting as themselves). The results have higher levels of education” (p.170). From the organizational are shown in Figure 1, where the dependent variable is the and financial streams of literature, Mitchell and Lusardi (2015) proportion of risky choices. It is easy to observe that the called for financial education in the workplace, in which they role shifting to “acting as a medical worker” largely reduced observed that many decisions, such as retirement saving and the probability of risk-seeking for both groups, irrespective of pension contribution rates, require a certain level of financial the order of the perspectives that they encountered. Trial 2 literacy. Both Lusardi (2008) and Oehler and Werner (2008) was under the negative frame, in which one group of subjects suggested that education could supplement structural pension (Group 1) received the Q3 scenario (imagining themselves arrangements such as automatic enrollment. In a recent study as the president) first and a second group first (Group 2) that directly investigated how framing effects could be used by received the Q2 scenario (imagining themselves as a medical HR teams in developing pension structure and communication worker). The results are shown below in Figure 2, where policies, Maloney and McCarthy (2016) suggested a significant the dependent variable remains the same. Comparing the role for personal education and financial literacy in achieving two manipulated roles, the results are similar to those of more desirable outcomes and improving the efficacy of such Trial 1. That is, there is no significant order effect actually decision-making processes. Arcidiacono (2011, p. 521) claimed incurred. that “: : :it is mainly those who have a lower income or a lower Materials and Procedure education level who are more likely to fall prey to framing.” In a similar vein, this study examines whether college students’ The current study slightly modified the Asian Disease Problem reasoning increases with a higher level of education, as expected developed by Tversky and Kahneman (1981) by manipulating the by the Chinese HE expansion policy. participants’ roles as “a medical worker” and “the president.” It is worth noting that how the participants would respond to such manipulation was somewhat unpredictable, so some assumptions Fan et al. (2015) provided supportive evidence regarding the effect of an HE expansion policy that works through a spillover channel on individual income. are provided below: Frontiers in Psychology | www.frontiersin.org 2 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 3 Fan Framing Effect in China ended with an odd digit, whereas Problem 2 was assigned to students whose numbers ended with an even digit. In the engineering class (the larger class), Problem 1 was assigned to students who sat in an odd column on the day of the testing, whereas Problem 2 was assigned to those who sat in the even columns. Both selection methods fit the definition of random assignment because the students received random student numbers upon admission, and they were free to sit wherever they wanted on the day of testing with no prior notification of the experiment. Q1: Imagine that China is preparing for the outbreak of an unusual disease that is expected to kill 6000 people. Two alternative programs to combat the disease have been proposed. FIGURE 1 | Counterbalanced orders (positive frame). Assume that the exact scientific estimates of the consequences of the programs are as follows: (Problem 1) If Program A is adopted, 2000 people will be saved. If Program B is adopted, there is a 1/3 probability that 6000 people will be saved and a 2/3 probability that no people will be saved. Which of the two programs would you favor? The second group was provided the cover story for Problem 1 but with different alternatives: (Problem 2) If Program C is adopted, 4000 people will die. If Program D is adopted, there is a 1/3 probability that FIGURE 2 | Counterbalanced orders (negative frame). nobody will die and a 2/3 probability that 6000 people will die. Which of the two programs would you favor? A1: “A medical worker” would provide a sense of The current study also added two new questions not professionalism, sympathy and caring in this life-and- previously used by Tversky and Kahneman (1981) by asking death problem that would make the participants more willing the students to imagine themselves in different roles for further to choose the probability of saving everyone over the definite investigative purposes. The Problems and Programs were the loss of some lives. same as Q1, so they are not listed again here. A2: “The president” would deliver a sense of power or a “big figure” feeling that may induce participants to consider the Q2: The problem you are facing is the same as above, problem in terms of the overall (even political) picture. but now imagine you are a medical worker. Which of the two programs would you favor? Students were assigned to Problems 1 or 2 based on their Q3: The problem you are facing is the same as student numbers or seats. Because different majors were involved above, but now imagine you are the president. Which of in the experiment, the group assignment methods differed the two programs would you favor? according to the class size. In the social science class (the smaller class), students were assigned Problem 1 if their student number RESULTS People might be concerned about how certain figures such as nurses are trained in China, and in turn whether the level of training would affect the results. In Main Results China, after finishing compulsory schooling (graduating from middle school), some students, often those who failed to gain acceptance to a higher-level school, For all three questions shown above, it is clear to see the two may choose a school of nursing and subsequently take a nursing job. Most such problems in each question are effectively identical. The only students would never receive higher education. Unlike many developed countries difference between them is that for Problem 1, the outcomes (e.g., the US), nursing education in China is still under-developed in terms of are framed in positive (described as the number of lives saved) structure, curriculum and faculty training, and the challenges are enormous for Chinese nursing education to meet international standards (Xu et al., 2000). Some terms and for Problem 2, the outcomes are framed in negative recent studies have indicated that although the scale of nursing education at three (described as the number of lives lost) terms. In the text and tables levels to enter nursing (baccalaureate degree, advanced diploma and secondary below, the questions are labeled ‘Q1,’ ‘Q2,’ and ‘Q3.’ The total diploma) has expanded rapidly in recent years, the current nursing students are still lacking in terms of their education and training levels (You et al., 2010). number of respondents for each problem is denoted by N, and Frontiers in Psychology | www.frontiersin.org 3 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 4 Fan Framing Effect in China the percentage of participants who chose each option is indicated The mean differences in risky choices among the three in brackets. scenarios using positive and negative frames are reported for the full sample in Table 1. The results show that there is no Q1 Scenario framing effect in Q1 (t D 1.63, p > 0.05, d D 0.30, 95% Problem 1 [N D 191]: confidence interval (CI) D [1.11, 1.68]) and a strong risk- seeking unidirectional framing effect in Q2 (tD2.94, p < 0.05, If Program A is adopted, 2000 people will be saved. [42%] dD 2.04, 95% CID [0.20, 3.77]). If Program B is adopted, there is a 1/3 probability that 6000 Notably, a mild bidirectional framing effect appears in Q3 people will be saved and a 2/3 probability that no one will be when the participants imagined themselves to be the president saved. [58%] (t D 2.28, p D 0.06, d D 0.94, 95% CI D [0.57, 2.39]), suggesting that the students’ choices were somewhat irrational No risk aversion emerged in this case. By contrast, the in that case because “acting as the president” is similar to participants were more risk-seeking in both the positive a daydream and is rather removed from reality. This result (tD4.42, p < 0.05) and negative (tD4.53, p < 0.05) framing occurred because the consequences of life-or-death decisions conditions. made by a medical worker or the president might be thought to be different from those made by the participants themselves in Q2 Scenario terms of the influence of acting in a certain role. This adaptive Problem 1 [N D 191]: information makes the subject’s risk proclivity irrational, and the If Program A is adopted, 2000 people will be saved. [45%] framing effect is thus more likely to be observed in this context. If Program B is adopted, there is a 1/3 probability that 6000 This result, called the perspective-specific risk preference, is people will be saved and a 2/3 probability that no one will be compatible with prior findings involving human reasoning (see saved. [55%] Gigerenzer and Hug, 1992). Problem 2 [N D 160]: Sex Differences To examine the heterogeneity effects, the sample was divided by If Program C is adopted, 4000 people will die. [28%] sex, and the corresponding results are reported in Table 2. As If Program D is adopted, there is a 1/3 probability that no one expected, the framing effects tended to vary by sex and remained will die and a 2/3 probability that 6000 people will die. [72%] role-related. There appears to be a significant sex difference for Notably, the participants’ choices changed to a certain extent all three manipulated roles under study. The male students did when they imagined themselves in the role of a medical worker. not show a decision bias for either being themselves or being a They were neither risk-averse nor risk-seeking in response to medical worker (tD1.24, p > 0.05, dD 0.36, 95% CID [2.31, Problem 1 (t D 1.21, p > 0.05); however, they continued to 1.66]) until their role was manipulated to that of the president, exhibit a strong risk-taking tendency in response to Problem 2 a role that is assumed to be more responsible and influential (tD17.07, p < 0.05). (tD3.27, p < 0.05, dD 2.5, 95% CID [0.57, 5.41]). Notably, this result shows a considerable bidirectional framing effect of Q3 Scenario risk-seeking, implying that the male students’ judgments were Problem 1 [N D 191]: greatly influenced and that there was an increase in ambiguity concerning the choice problem when they imagined themselves If Program A is adopted, 2000 people will be saved. [58%] to be president. This result might be explained as it was in Wang’s If Program B is adopted, there is a 1/3 probability that 6000 (1996b) study: “If subjects do not really care whether they choose people will be saved and a 2/3 probability that no one will be a sure outcome or a gamble, then a minor variation in wording or saved. [42%] phrasing may greatly influence their choices” (p. 11). Problem 2 [N D 160]: It would also be interesting to divide the sample by major. However, the sample size difference (304 vs. 45) is so extreme that there can be little confidence in the If Program C is adopted, 4000 people will die. [38%] results. Increasing the sample size of social science majors in a future experiment If Program D is adopted, there is a 1/3 probability that no one would allow for further examination of this issue. will die and a 2/3 probability that 6000 people will die. [62%] When the participants imagined themselves in the role of TABLE 1 | Mean difference in Q1–Q3 by frame. president, the results were similar to those found in previous studies; that is, there was a general preference reversal from N D 351 M positive M negative M positive–M t p a predominantly risk-averse choice under a positive frame negative (framing effect) (t D 11.62, p < 0.05) to a tiny risk-seeking preference under a negative frame (tD1.89, p > 0.05). Q1 0.585 0.692 0.107 1.63 0.147 Q2 0.548 0.721 0.173 2.94 0.022 Notably, the different assignments resulted in relatively large differences between Q3 0.417 0.615 0.198 2.28 0.057 the sample sizes of the two groups (191 students for Problem 1 vs. 160 students for Problem 2). p < 0.05. Frontiers in Psychology | www.frontiersin.org 4 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 5 Fan Framing Effect in China TABLE 2 | Mean difference in Q1–Q3 by frame and sex. to the framing manipulations, which is consistent with the previous literature—Peters and Levin (2008, p, 441) found that Male “individuals lower in numeracy were influenced more than N D 122 M positive M negative M positive–M t p those higher in numeracy by the frame in the risky-choice negative scenarios.” Furthermore, the present study can explore the (framing effect) changes being affected in the current Chinese higher education system by comparing this result to that of a similar experiment Q1 0.552 0.671 0.119 1.24 0.303 conducted with Chinese military students who had received less Q2 0.674 0.728 0.054 1.24 0.303 mathematical training during college. This group can serve as Q3 0.4 0.663 0.263 3.27 0.047 a control group in this situation. In that work, Zhang et al. Female (2008) found that the military participants were influenced by both positive and negative framing. It should be noted that there N D 229 M positive M negative M positive–M t p negative might be some symmetrical differences between participants (framing effect) with different majors: recent research has reported a relatively low return for engineering higher education in China (Fan and Q1 0.599 0.817 0.219 12.78 0.001 Zhang, 2015). All of these issues call for further study that links Q2 0.492 0.677 0.185 12.61 0.001 the education policy to individual cognition and decision-making Q3 0.425 0.324 0.101 1.19 0.321 through this channel. p < 0.05. Second, the framing effects may be attenuated or offset because of the diverse composition of the entire sample. In By contrast, female students showed a significant framing other words, human risk preferences and choice strategies are effect when both being themselves and a medical worker sensitive to a decision-maker’s biological conditions (such as age (t D 12.78, p < 0.05, d D 5.66, 95% CI D [0.35, 11.15]; and sex) as well as to the adaptive values inherent in the choice t D 12.61, p < 0.05, d D 6.1, 95% CI D [0.45, 11.99]). options (Wang, 1996a). However, a strong unidirectional framing In particular, decisions were highly biased by the bidirectional effect was found in Q2 (when participants imagined themselves framing effects of risk-seeking when they imagined themselves as medical workers) because a decision-maker may become in the role of a medical worker, rather than as the president more risk and variance seeking to maximize the probability of (t D 1.19, p > 0.05, d D 0.14, 95% CI D [1.84, 2.09]). This achieving a goal when the mean expected value is below her result implies that there is a sex difference in the sensitivity to minimum requirement (Wang, 1996b). In this life-and-death adaptive information embodied in the decision problem. Men problem, the minimum requirement tends to be higher than the tend to be more influenced by a word that reflects certain mean expected value of the choice option—the probability of masculine traits, such as aggressiveness (as is the case when saving everyone is preferred over the definite loss of some lives. they are acting in the role of president), whereas women pay With regard to sex differences, the present research supports more attention to words that represent sympathy and caring a number of existing studies. Hasseldine and Hite (2003) (as is the case when they are acting in the role of a medical found no evidence of a main effect for framing objectively worker). equivalent information; however, a significant frame through a gender interaction effect was documented. They reported that “females were more compliant in response to a positively framed DISCUSSION persuasive communication than they were to the negatively framed message, even though the information in each message The discussion is twofold. First, the main results are interpreted. was objectively equivalent” (pp. 528–529). Saad and Gill (2014) Second, the heterogeneity effects are examined. showed that women have greater sensitivity to negatively framed As discussed above, no framing effect was found in Q1, information than men do. This study supports the latter finding but a strong unidirectional framing effect was found in Q2. by showing that more than 80% of the female students selected How should one interpret this distinction? There are at least the risky choice in the negative frame in Q1. Additionally, two reasons why no framing effect was found in Q1. First, for female participants only, a conversion from unidirectional Chinese students generally have strong computing skills that are framing effects to bidirectional framing effects associated with developed in the primary educational system ; the respondents the perspective shift from oneself to a medical worker was in this study were particularly well trained in college economics observed. This finding is consistent with Huang and Wang and finance courses. These features make their risk preferences (2010), who argued that the prior design that focused on overall somewhat unambiguous such that they are more immune sex differences might be misleading because sex differences in framing effects might vary across different domains. Overall, the Shanghai has topped two consecutive rounds of the Programme for International present study echoes and emphasizes the relationship between Student Assessment (PISA) tests in reading, mathematics and science. The Organization for Economic Cooperation and Development (OECD) administers decision-making and gender, which is an interesting topic that these tests to assess how well 15-year-olds have acquired the knowledge and skills is growing in popularity and receiving increased attention in needed to fully participate in knowledge-driven societies (OECD, 2009, 2012). In various fields (e.g., Mendelberg and Karpowitz, 2016; Alwang addition, a direct measure of numeracy has been designed and will be used in the next experiments. et al., 2017). Frontiers in Psychology | www.frontiersin.org 5 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 6 Fan Framing Effect in China information embodied in the decision problem. Notably, this CONCLUSION finding coincides with the doctrine of the Chinese philosopher Conventional empirical evidence seems to lead to the conclusion and educator Confucius, one of whose well-known dictums that there are no substantial differences in the rationality of is “teach students in accordance with their aptitude.” This performance when considering variables such as gender, age, study may assist in efficiently guiding further policy reforms education or social status, which is partly due to a lack of in this direction. In addition, the findings also imply that the standardization and regularity in the methodologies used (e.g., subject’s perspective is an important factor in the decision- Gigerenzer and Murrey, 1987; Hertwig and Ortmann, 2001, making process. 2005). This experimental study examined the framing effect Finally, one potential issue is that the participants may have of Chinese college students with the aim of evaluating the only provided their estimation of how “certain figures” would effectiveness of HE reform from a different perspective. The respond to each scenario. The participants may lack adequate results show that the participants’ preferences were basically knowledge of these manipulated roles, which might undermine unbiased and at least partially driven by the high numeracy the results. In this regard, more preliminary work could be skills the participants had accumulated through higher education, conducted in an attempt to improve the research design in the which demonstrates the weakness of classical framing effects. future. Thus, this study offers new evidence on the relationship between education and behavior in China. Furthermore, this study sheds light on how such an analysis might help to elucidate evaluations AUTHOR CONTRIBUTIONS of the effects of China’s HE policy, under which more young people have access to higher education and receive better training WF designed and carried out experiments, as well as analyzed the in thinking and reasoning skills. China’s educational system results and wrote the manuscript. has clearly been expanding on a sustainable developmental path. When the student responses were sensitively manipulated, FUNDING however, dividing the sample by sex revealed a tendency for them to be influenced by framing effects. A significant gender I gratefully acknowledge financial supports from the Priority difference emerged in this context, which calls for more research Academic Program Development of Jiangsu Higher Education to deeply determine whether the framing effects are susceptible Institutions (PAPD); SRF for ROCS, SEM; Key Program of to individual differences and sheds light on improving teaching Humanity and Social Science foundation, State Education output. More consideration should therefore be given to teaching Ministry 16JJD840008; Nanjing University of Finance and styles and material as well as students’ aptitudes in light Economics Pre-research fund YYJ201413 and “Young Scholar” of the sex differences revealed in sensitivity to the adaptive Program. Harris, P. (2008). Designing and Reporting Experiments in Psychology, 3rd Edn. REFERENCES Maidenhead: Open University Press. Alwang, J., Larochelle, C., and Barrera, V. (2017). Farm decision making and Hasseldine, J., and Hite, P. A. (2003). Framing, gender and tax compliance. J. Econ. gender: results from a randomized experiment in Ecuador. World Dev. 92, Psychol. 24, 517–533. doi: 10.1016/S0167- 4870(02)00209- X 117–129. doi: 10.1016/j.worlddev.2016.11.015 Hertwig, R., and Ortmann, A. (2001). Experimental practices in economics: a Anderson, C. L., Reynolds, T. W., and Gugerty, M. K. (2017). Husband and methodological challenge for psychologists? Behav. Brain Sci 24, 383–403; wife perspectives on farm household decision-making authority and evidence discussion 403–451. doi: 10.1037/e683322011- 032 on intra-household accord in rural Tanzania. World Dev. 90, 169–183. Hertwig, R., and Ortmann, A. (2005). “The cognitive illusion controversy: a doi: 10.1016/j.worlddev.2016.09.005 methodological debate in disguise that matters to economists,” in Experimental Arcidiacono, D. (2011). Consumer rationality in a multidisciplinary perspective. Business Research, Vol. 3, eds R. Zwick and A. Rapaport (New York, NY: J. Socioecon. 40, 516–522. doi: 10.1016/j.socec.2011.01.003 Springer), 113–130. doi: 10.1007/0- 387- 24244- 9_5 Chambaere, K., Rietjens, J. A., Cohen, J., Pardon, K., Deschepper, R., Pasman, Huang, Y., and Wang, L. (2010). Sex differences in framing effects across task H. R. W., et al. (2013). Is educational attainment related to end-of-life decision- domain. Personal. Individ. Dif. 48, 649–653. doi: 10.1016/j.paid.2010.01.005 making? A large post-mortem survey in Belgium. BMC Public Health 13:1055. Kidd, A. M., Monz, C., D’Antonio, A., Manning, R. E., Reigner, N., Goonan, doi: 10.1186/1471- 2458- 13- 1055 K. A., et al. (2015). The effect of minimum impact education on visitor spatial Fan, J., and Zhang, C. (2015). A study of the rate of return to higher engineering behavior in parks and protected areas: an experimental investigation using GPS- education in China. Int. J. Educ. Dev. 42, 106–114. doi: 10.1016/j.ijedudev.2015. based tracking. J. Environ. Manage. 162, 53–62. doi: 10.1016/j.jenvman.2015. 04.010 07.007 Fan, W., Ma, Y. Y., and Wang, L. M. (2015). Do we need more public investment in Levin, I. P., Gaeth, G. J., Schreiber, J., and Lauriola, M. (2002). A new look higher education? Estimating the external returns to higher education in China. at framing effects: distribution of effect sizes, individual differences, and Asian Econ. Pap. 14, 88–104. doi: 10.1162/ASEP_a_00380 independence of types of effects. Organ. Behav. Hum. Decis. Process. 88, Gigerenzer, G., and Hug, K. (1992). Domain-specific reasoning: social contracts, 411–429. doi: 10.1006/obhd.2001.2983 cheating, and perspective change. Cognition 43, 127–171. doi: 10.1016/0010- Lusardi, A. (2008). Household Saving Behavior: The Role of Literacy, Information 0277(92)90060- U and Financial Education Programs. Hanover, PA: Dartmouth College. Gigerenzer, G., and Murrey, D. J. (1987). Cognition as Intuitive Statistics. Hillsdale, Maloney, M., and McCarthy, A. (2016). Understanding pension communications NJ: Erlbaum. at the organizational level: insights from bounded rationality theory & Frontiers in Psychology | www.frontiersin.org 6 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 7 Fan Framing Effect in China implications for HRM. Hum. Res. Manage. Rev. 27, 338–352. doi: 10.1016/j. Tversky, A., and Kahneman, D. (1986). Rational choice and the framing of hrmr.2016.08.001 decisions. J. Bus. 59, S251–S278. doi: 10.1086/296365 Mendelberg, T., and Karpowitz, C. F. (2016). Women’s authority in political Wang, X. T. (1996a). Evolutionary hypotheses of risk-sensitive choice: age decision-making groups. Leadersh. Q. 27, 487–503. doi: 10.1016/j.leaqua.2015. differences and perspective change. Ethol. Sociobiol. 17, 1–15. doi: 10.1016/ 11.005 0162- 3095(95)00103- 4 Mitchell, O., and Lusardi, A. (2015). Financial Literacy and Economic Outcomes: Wang, X. T. (1996b). Framing effects: dynamics and task domains. Organ. Evidence and Policy Implication. Working Paper. Philadelphia, PA: University of Behav. Hum. Decis. Process. 68, 145–157. doi: 10.1006/obhd.1996. Pennsylvania. 0095 Nafría, J. M. D., Cendón, J. A., and Alonso, L. P. (2015). Building up eParticipatory Xu, Y., Xu, Z., and Zhang, J. (2000). The nursing education system in the People’s decision-making from the local to the global scale. Study case at the European Republic of China: evolution, structure and reform. Int. Nurs. Rev. 47, 207–217. higher education area. Comput. Hum. Behav. 47, 26–41. doi: 10.1016/j.cnb.2014. doi: 10.1046/j.1466- 7657.2000.00027 09.004 You, L. M., Wan, L. H., Liu, K., Zhu, X. W., Zheng, J., and Luo, Z. M. (2010). OECD (2009). OECD Reports on PISA Results. Available at: http://www.oecd- Education capacity of nursing schools and the needs to nurses in China. Nurs. ilibrary.org/education/pisa_19963777 Manage. 10, 46–50. OECD (2012). OECD Reports on PISA Results. Available at: http://www.oecd- Zhang, Y. L., and Miao, D. M. (2008). Social cues and framing effects in risky ilibrary.org/education/pisa_19963777 decisions among Chinese military students. Asian J. Soc. Psychol. 11, 241–246. Oehler, A., and Werner, C. (2008). Saving for retirement—a case for financial doi: 10.1111/j.1467- 839X.2008.00263.x education in Germany and UK? An economic perspective. J. Con. Policy 31, Zhang, Y. L., Xiao, L. J., Ma, Y., and Miao, D. M. (2008). Effect of framing on risky 253–283. doi: 10.1007/s10603- 008- 9074- 5 choice: one case study in China. Soc. Behav. Pers. 36, 651–658. doi: 10.2224/sbp. Peters, E., and Levin, I. P. (2008). Dissecting the risky-choice framing effect: 2008.36.5.651 numeracy as an individual-difference factor in weighting risky and riskless options. Judgm. Decis. Mak. 3, 435–448. doi: 10.2307/1914185 Conflict of Interest Statement: The author declares that the research was Petroman, C., Bidireac, I. C., Petroman, I., Sucan, M., Marin, D., Turc, B., et al. conducted in the absence of any commercial or financial relationships that could (2015). The impact of education on the behaviour of the consumer of animal be construed as a potential conflict of interest. origin food products. Procedia Soc. Behav. Sci. 190, 429–433. doi: 10.1016/j. sbspro.2015.05.021 Copyright © 2017 Fan. This is an open-access article distributed under the terms Saad, G., and Gill, T. (2014). The framing effect when evaluating prospective mates: of the Creative Commons Attribution License (CC BY). The use, distribution or an adaptationist perspective. Evol. Hum. Behav. 35, 184–192. doi: 10.1016/j. reproduction in other forums is permitted, provided the original author(s) or licensor evolhumbehav.2014.01.002 are credited and that the original publication in this journal is cited, in accordance Tversky, A., and Kahneman, D. (1981). The framing of decisions and the with accepted academic practice. No use, distribution or reproduction is permitted psychology of choice. Science 211, 453–458. doi: 10.1126/science.7455683 which does not comply with these terms. Frontiers in Psychology | www.frontiersin.org 7 May 2017 | Volume 8 | Article 744 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Frontiers in Psychology Pubmed Central

Education and Decision-Making: An Experimental Study on the Framing Effect in China

Frontiers in Psychology , Volume 8 – May 11, 2017

Loading next page...
 
/lp/pubmed-central/education-and-decision-making-an-experimental-study-on-the-framing-y9yJ0T2yiE

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
Copyright
Copyright © 2017 Fan.
ISSN
1664-1078
eISSN
1664-1078
DOI
10.3389/fpsyg.2017.00744
Publisher site
See Article on Publisher Site

Abstract

fpsyg-08-00744 May 9, 2017 Time: 15:44 # 1 ORIGINAL RESEARCH published: 11 May 2017 doi: 10.3389/fpsyg.2017.00744 Education and Decision-Making: An Experimental Study on the Framing Effect in China 1,2 Wen Fan * 1 2 School of Public Administration, Nanjing University of Finance and Economics, Nanjing, China, Center for Social Security Study, Wuhan University, Wuhan, China China’s higher education expansion policy has been in effect for almost two decades. Under this policy, a growing number of youths have gained access to higher education, which aims to train students to be more rational. This study examines human rationality at a Chinese college through an experiment based on the risky-choice framing effect. The basic results show no classical framing effect with regard to individual decisions for the entire sample in a benchmark setting. However, when the participants’ roles were manipulated and subsamples were investigated, a significant framing effect was found that appeared to be role-related and that varied by sex. These results help to elucidate evaluations of the effects of China’s higher education policy and may assist in guiding further policy reforms. Keywords: framing effect, decision making, heterogeneity, higher education policy, China Edited by: Michael S. Dempsey, INTRODUCTION Boston University, USA Reviewed by: People make many decisions and judgments in daily life, the explanations for and predictions of Renata Melinda Heilman, which are often based on the assumption of human rationality. However, as Tversky and Kahneman Babes-Bolyai ¸ University, Romania (1986) stated, “Alternative descriptions of decision problems often lead to different preferences, Azizuddin Khan, contrary to the principle of invariance that underlies the rational theory of choice. Violations of Indian Institute of Technology this theory are traced to the rules that govern the framing of decisions and to the psychophysical Bombay, India principles of evaluation embodied in prospect theory” (p. S251). This decision bias is well known as *Correspondence: a risky-choice framing effect, in which choices involving gains are frequently risk-averse, whereas Wen Fan choices involving losses are frequently evidence of risk-taking. The framing effect is commonly nonpareilian@gmail.com considered to be one of the most severe violations of normative utility axioms and is therefore a strong indicator of irrationality. Thus, individuals’ preferences will reverse or shift (bidirectional Specialty section: This article was submitted to or unidirectional framing effects, respectively) when the same problem is framed in different ways. Educational Psychology, Classic bidirectional framing effects frequently lead to irrational reversals in risk preferences under a section of the journal different framing conditions. However, as Wang (1996b) defined the term, “[the] unidirectional Frontiers in Psychology framing effect involves no preference reversal but a shift to a more extreme risk preference: : : if the Received: 29 August 2016 predominant preference is unidirectionally risk-averse under both framing conditions, it is even Accepted: 24 April 2017 more risk-averse when positively, as opposed to negatively, framed. Similarly, if the predominant Published: 11 May 2017 preference is unidirectionally risk-seeking under both framing conditions, it is even more so under Citation: a negative frame” (p. 5). Fan W (2017) Education This study investigates human rationality through an experiment that examines the risky- and Decision-Making: An choice framing effect. The participants were undergraduate students from two different majors Experimental Study on the Framing at a Chinese college. Many studies have been conducted with US college students (e.g., Tversky Effect in China. Front. Psychol. 8:744. doi: 10.3389/fpsyg.2017.00744 and Kahneman, 1981; Levin et al., 2002). However, recent studies in the Chinese context have been Frontiers in Psychology | www.frontiersin.org 1 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 2 Fan Framing Effect in China limited by small sample sizes or the lack of a detailed analysis. MATERIALS AND METHODS For example, Zhang and Miao (2008) conducted similar lab experiments, but the authors focused on the experiment per Participants and Design se. Zhang et al. (2008) included both military and civilian The participants included 351 students (65% female, female mean students, but the authors did not explore heterogeneity effects. age D 21.6, male mean age D 21.8) studying in two majors Compared to the sample sizes of similar experiments conducted (mean age D 20 for social science and mean age D 22 for in developed countries, this study’s larger sample size (N D 351) engineering) at a 4-year college in eastern China; the students yields more power for testing the reliability of the framing received class credit in exchange for their participation in the effects. In addition, recruiting participants from diverse majors experiment. The responses were anonymous, and the instructions allowed a further exploration of the different individual factors specified that although there was no “correct” answer to the that have been employed as determinants (e.g., self-esteem, problem, careful thinking would be highly appreciated. The numeracy, and biological conditions). Thus, this lab experiment students were not permitted to speak with one another during at a Chinese college contributes new evidence to the literature the experiment. All questions were presented in written form, by allowing identification of the perspective- and sex-dependent and the split-ballot questionnaire was administered in two choice patterns that reveal that human cognitive mechanisms classrooms. The study was approved by the research ethics are sensitive to the internal biological status of the information- committee of Nanjing University of Finance and Economics, and processing organism. conforms to the ethical principles of the Declaration of Helsinki (World Medical Association, 2013). Before administering the In addition, as the Chinese Higher Education (HE) expansion policy has been in effect for almost two decades, the public experiment, one of the instructors translated the principle of is highly interested in the linkage between this nationwide informed consent according to Standard 8.02 of the American education policy and the social and economic changes it has Psychological Association’s (APA) new ethics code into Chinese brought to Chinese life and society. There is substantial literature and explained it to all of the participants. The experiment had on how education in different disciplines relates to people’s a two-way mixed design in which an alpha level of 0.05 was choices (see Chambaere et al., 2013, for end-of-life research; Kidd used for all t-tests. The related measures independent variable et al., 2015, for pro-environment behavior analysis; Nafría et al., was the within-participants manipulated role (general, medical 2015, for an eParticipatory decision-making study; and Petroman worker, and president). The unrelated measures independent et al., 2015, for food consumption). variable was the between-participants decision-making domain Most importantly, there is an increasing body of literature (gain and loss). Following Harris (2008), a pilot test was showing that individual rationality or decision-making power administered on a randomly selected smaller group that allowed appears to be positively related to education level. Anderson the order of roles to be counterbalanced such that their order et al. (2017) found strong evidence that a husband’s allocation of presentation varied among the participants. Trial 1 was of decision-making authority to his wife varies according to his under the positive frame, in which one group of subjects wife’s age and education. The authors stated that “on average, (Group 1) first received the Q2 scenario (imagining themselves intra-household accord over which spouse holds decision- as a medical worker) and a second group (Group 2) first making authority is more likely in households where women received the Q1 scenario (just acting as themselves). The results have higher levels of education” (p.170). From the organizational are shown in Figure 1, where the dependent variable is the and financial streams of literature, Mitchell and Lusardi (2015) proportion of risky choices. It is easy to observe that the called for financial education in the workplace, in which they role shifting to “acting as a medical worker” largely reduced observed that many decisions, such as retirement saving and the probability of risk-seeking for both groups, irrespective of pension contribution rates, require a certain level of financial the order of the perspectives that they encountered. Trial 2 literacy. Both Lusardi (2008) and Oehler and Werner (2008) was under the negative frame, in which one group of subjects suggested that education could supplement structural pension (Group 1) received the Q3 scenario (imagining themselves arrangements such as automatic enrollment. In a recent study as the president) first and a second group first (Group 2) that directly investigated how framing effects could be used by received the Q2 scenario (imagining themselves as a medical HR teams in developing pension structure and communication worker). The results are shown below in Figure 2, where policies, Maloney and McCarthy (2016) suggested a significant the dependent variable remains the same. Comparing the role for personal education and financial literacy in achieving two manipulated roles, the results are similar to those of more desirable outcomes and improving the efficacy of such Trial 1. That is, there is no significant order effect actually decision-making processes. Arcidiacono (2011, p. 521) claimed incurred. that “: : :it is mainly those who have a lower income or a lower Materials and Procedure education level who are more likely to fall prey to framing.” In a similar vein, this study examines whether college students’ The current study slightly modified the Asian Disease Problem reasoning increases with a higher level of education, as expected developed by Tversky and Kahneman (1981) by manipulating the by the Chinese HE expansion policy. participants’ roles as “a medical worker” and “the president.” It is worth noting that how the participants would respond to such manipulation was somewhat unpredictable, so some assumptions Fan et al. (2015) provided supportive evidence regarding the effect of an HE expansion policy that works through a spillover channel on individual income. are provided below: Frontiers in Psychology | www.frontiersin.org 2 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 3 Fan Framing Effect in China ended with an odd digit, whereas Problem 2 was assigned to students whose numbers ended with an even digit. In the engineering class (the larger class), Problem 1 was assigned to students who sat in an odd column on the day of the testing, whereas Problem 2 was assigned to those who sat in the even columns. Both selection methods fit the definition of random assignment because the students received random student numbers upon admission, and they were free to sit wherever they wanted on the day of testing with no prior notification of the experiment. Q1: Imagine that China is preparing for the outbreak of an unusual disease that is expected to kill 6000 people. Two alternative programs to combat the disease have been proposed. FIGURE 1 | Counterbalanced orders (positive frame). Assume that the exact scientific estimates of the consequences of the programs are as follows: (Problem 1) If Program A is adopted, 2000 people will be saved. If Program B is adopted, there is a 1/3 probability that 6000 people will be saved and a 2/3 probability that no people will be saved. Which of the two programs would you favor? The second group was provided the cover story for Problem 1 but with different alternatives: (Problem 2) If Program C is adopted, 4000 people will die. If Program D is adopted, there is a 1/3 probability that FIGURE 2 | Counterbalanced orders (negative frame). nobody will die and a 2/3 probability that 6000 people will die. Which of the two programs would you favor? A1: “A medical worker” would provide a sense of The current study also added two new questions not professionalism, sympathy and caring in this life-and- previously used by Tversky and Kahneman (1981) by asking death problem that would make the participants more willing the students to imagine themselves in different roles for further to choose the probability of saving everyone over the definite investigative purposes. The Problems and Programs were the loss of some lives. same as Q1, so they are not listed again here. A2: “The president” would deliver a sense of power or a “big figure” feeling that may induce participants to consider the Q2: The problem you are facing is the same as above, problem in terms of the overall (even political) picture. but now imagine you are a medical worker. Which of the two programs would you favor? Students were assigned to Problems 1 or 2 based on their Q3: The problem you are facing is the same as student numbers or seats. Because different majors were involved above, but now imagine you are the president. Which of in the experiment, the group assignment methods differed the two programs would you favor? according to the class size. In the social science class (the smaller class), students were assigned Problem 1 if their student number RESULTS People might be concerned about how certain figures such as nurses are trained in China, and in turn whether the level of training would affect the results. In Main Results China, after finishing compulsory schooling (graduating from middle school), some students, often those who failed to gain acceptance to a higher-level school, For all three questions shown above, it is clear to see the two may choose a school of nursing and subsequently take a nursing job. Most such problems in each question are effectively identical. The only students would never receive higher education. Unlike many developed countries difference between them is that for Problem 1, the outcomes (e.g., the US), nursing education in China is still under-developed in terms of are framed in positive (described as the number of lives saved) structure, curriculum and faculty training, and the challenges are enormous for Chinese nursing education to meet international standards (Xu et al., 2000). Some terms and for Problem 2, the outcomes are framed in negative recent studies have indicated that although the scale of nursing education at three (described as the number of lives lost) terms. In the text and tables levels to enter nursing (baccalaureate degree, advanced diploma and secondary below, the questions are labeled ‘Q1,’ ‘Q2,’ and ‘Q3.’ The total diploma) has expanded rapidly in recent years, the current nursing students are still lacking in terms of their education and training levels (You et al., 2010). number of respondents for each problem is denoted by N, and Frontiers in Psychology | www.frontiersin.org 3 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 4 Fan Framing Effect in China the percentage of participants who chose each option is indicated The mean differences in risky choices among the three in brackets. scenarios using positive and negative frames are reported for the full sample in Table 1. The results show that there is no Q1 Scenario framing effect in Q1 (t D 1.63, p > 0.05, d D 0.30, 95% Problem 1 [N D 191]: confidence interval (CI) D [1.11, 1.68]) and a strong risk- seeking unidirectional framing effect in Q2 (tD2.94, p < 0.05, If Program A is adopted, 2000 people will be saved. [42%] dD 2.04, 95% CID [0.20, 3.77]). If Program B is adopted, there is a 1/3 probability that 6000 Notably, a mild bidirectional framing effect appears in Q3 people will be saved and a 2/3 probability that no one will be when the participants imagined themselves to be the president saved. [58%] (t D 2.28, p D 0.06, d D 0.94, 95% CI D [0.57, 2.39]), suggesting that the students’ choices were somewhat irrational No risk aversion emerged in this case. By contrast, the in that case because “acting as the president” is similar to participants were more risk-seeking in both the positive a daydream and is rather removed from reality. This result (tD4.42, p < 0.05) and negative (tD4.53, p < 0.05) framing occurred because the consequences of life-or-death decisions conditions. made by a medical worker or the president might be thought to be different from those made by the participants themselves in Q2 Scenario terms of the influence of acting in a certain role. This adaptive Problem 1 [N D 191]: information makes the subject’s risk proclivity irrational, and the If Program A is adopted, 2000 people will be saved. [45%] framing effect is thus more likely to be observed in this context. If Program B is adopted, there is a 1/3 probability that 6000 This result, called the perspective-specific risk preference, is people will be saved and a 2/3 probability that no one will be compatible with prior findings involving human reasoning (see saved. [55%] Gigerenzer and Hug, 1992). Problem 2 [N D 160]: Sex Differences To examine the heterogeneity effects, the sample was divided by If Program C is adopted, 4000 people will die. [28%] sex, and the corresponding results are reported in Table 2. As If Program D is adopted, there is a 1/3 probability that no one expected, the framing effects tended to vary by sex and remained will die and a 2/3 probability that 6000 people will die. [72%] role-related. There appears to be a significant sex difference for Notably, the participants’ choices changed to a certain extent all three manipulated roles under study. The male students did when they imagined themselves in the role of a medical worker. not show a decision bias for either being themselves or being a They were neither risk-averse nor risk-seeking in response to medical worker (tD1.24, p > 0.05, dD 0.36, 95% CID [2.31, Problem 1 (t D 1.21, p > 0.05); however, they continued to 1.66]) until their role was manipulated to that of the president, exhibit a strong risk-taking tendency in response to Problem 2 a role that is assumed to be more responsible and influential (tD17.07, p < 0.05). (tD3.27, p < 0.05, dD 2.5, 95% CID [0.57, 5.41]). Notably, this result shows a considerable bidirectional framing effect of Q3 Scenario risk-seeking, implying that the male students’ judgments were Problem 1 [N D 191]: greatly influenced and that there was an increase in ambiguity concerning the choice problem when they imagined themselves If Program A is adopted, 2000 people will be saved. [58%] to be president. This result might be explained as it was in Wang’s If Program B is adopted, there is a 1/3 probability that 6000 (1996b) study: “If subjects do not really care whether they choose people will be saved and a 2/3 probability that no one will be a sure outcome or a gamble, then a minor variation in wording or saved. [42%] phrasing may greatly influence their choices” (p. 11). Problem 2 [N D 160]: It would also be interesting to divide the sample by major. However, the sample size difference (304 vs. 45) is so extreme that there can be little confidence in the If Program C is adopted, 4000 people will die. [38%] results. Increasing the sample size of social science majors in a future experiment If Program D is adopted, there is a 1/3 probability that no one would allow for further examination of this issue. will die and a 2/3 probability that 6000 people will die. [62%] When the participants imagined themselves in the role of TABLE 1 | Mean difference in Q1–Q3 by frame. president, the results were similar to those found in previous studies; that is, there was a general preference reversal from N D 351 M positive M negative M positive–M t p a predominantly risk-averse choice under a positive frame negative (framing effect) (t D 11.62, p < 0.05) to a tiny risk-seeking preference under a negative frame (tD1.89, p > 0.05). Q1 0.585 0.692 0.107 1.63 0.147 Q2 0.548 0.721 0.173 2.94 0.022 Notably, the different assignments resulted in relatively large differences between Q3 0.417 0.615 0.198 2.28 0.057 the sample sizes of the two groups (191 students for Problem 1 vs. 160 students for Problem 2). p < 0.05. Frontiers in Psychology | www.frontiersin.org 4 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 5 Fan Framing Effect in China TABLE 2 | Mean difference in Q1–Q3 by frame and sex. to the framing manipulations, which is consistent with the previous literature—Peters and Levin (2008, p, 441) found that Male “individuals lower in numeracy were influenced more than N D 122 M positive M negative M positive–M t p those higher in numeracy by the frame in the risky-choice negative scenarios.” Furthermore, the present study can explore the (framing effect) changes being affected in the current Chinese higher education system by comparing this result to that of a similar experiment Q1 0.552 0.671 0.119 1.24 0.303 conducted with Chinese military students who had received less Q2 0.674 0.728 0.054 1.24 0.303 mathematical training during college. This group can serve as Q3 0.4 0.663 0.263 3.27 0.047 a control group in this situation. In that work, Zhang et al. Female (2008) found that the military participants were influenced by both positive and negative framing. It should be noted that there N D 229 M positive M negative M positive–M t p negative might be some symmetrical differences between participants (framing effect) with different majors: recent research has reported a relatively low return for engineering higher education in China (Fan and Q1 0.599 0.817 0.219 12.78 0.001 Zhang, 2015). All of these issues call for further study that links Q2 0.492 0.677 0.185 12.61 0.001 the education policy to individual cognition and decision-making Q3 0.425 0.324 0.101 1.19 0.321 through this channel. p < 0.05. Second, the framing effects may be attenuated or offset because of the diverse composition of the entire sample. In By contrast, female students showed a significant framing other words, human risk preferences and choice strategies are effect when both being themselves and a medical worker sensitive to a decision-maker’s biological conditions (such as age (t D 12.78, p < 0.05, d D 5.66, 95% CI D [0.35, 11.15]; and sex) as well as to the adaptive values inherent in the choice t D 12.61, p < 0.05, d D 6.1, 95% CI D [0.45, 11.99]). options (Wang, 1996a). However, a strong unidirectional framing In particular, decisions were highly biased by the bidirectional effect was found in Q2 (when participants imagined themselves framing effects of risk-seeking when they imagined themselves as medical workers) because a decision-maker may become in the role of a medical worker, rather than as the president more risk and variance seeking to maximize the probability of (t D 1.19, p > 0.05, d D 0.14, 95% CI D [1.84, 2.09]). This achieving a goal when the mean expected value is below her result implies that there is a sex difference in the sensitivity to minimum requirement (Wang, 1996b). In this life-and-death adaptive information embodied in the decision problem. Men problem, the minimum requirement tends to be higher than the tend to be more influenced by a word that reflects certain mean expected value of the choice option—the probability of masculine traits, such as aggressiveness (as is the case when saving everyone is preferred over the definite loss of some lives. they are acting in the role of president), whereas women pay With regard to sex differences, the present research supports more attention to words that represent sympathy and caring a number of existing studies. Hasseldine and Hite (2003) (as is the case when they are acting in the role of a medical found no evidence of a main effect for framing objectively worker). equivalent information; however, a significant frame through a gender interaction effect was documented. They reported that “females were more compliant in response to a positively framed DISCUSSION persuasive communication than they were to the negatively framed message, even though the information in each message The discussion is twofold. First, the main results are interpreted. was objectively equivalent” (pp. 528–529). Saad and Gill (2014) Second, the heterogeneity effects are examined. showed that women have greater sensitivity to negatively framed As discussed above, no framing effect was found in Q1, information than men do. This study supports the latter finding but a strong unidirectional framing effect was found in Q2. by showing that more than 80% of the female students selected How should one interpret this distinction? There are at least the risky choice in the negative frame in Q1. Additionally, two reasons why no framing effect was found in Q1. First, for female participants only, a conversion from unidirectional Chinese students generally have strong computing skills that are framing effects to bidirectional framing effects associated with developed in the primary educational system ; the respondents the perspective shift from oneself to a medical worker was in this study were particularly well trained in college economics observed. This finding is consistent with Huang and Wang and finance courses. These features make their risk preferences (2010), who argued that the prior design that focused on overall somewhat unambiguous such that they are more immune sex differences might be misleading because sex differences in framing effects might vary across different domains. Overall, the Shanghai has topped two consecutive rounds of the Programme for International present study echoes and emphasizes the relationship between Student Assessment (PISA) tests in reading, mathematics and science. The Organization for Economic Cooperation and Development (OECD) administers decision-making and gender, which is an interesting topic that these tests to assess how well 15-year-olds have acquired the knowledge and skills is growing in popularity and receiving increased attention in needed to fully participate in knowledge-driven societies (OECD, 2009, 2012). In various fields (e.g., Mendelberg and Karpowitz, 2016; Alwang addition, a direct measure of numeracy has been designed and will be used in the next experiments. et al., 2017). Frontiers in Psychology | www.frontiersin.org 5 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 6 Fan Framing Effect in China information embodied in the decision problem. Notably, this CONCLUSION finding coincides with the doctrine of the Chinese philosopher Conventional empirical evidence seems to lead to the conclusion and educator Confucius, one of whose well-known dictums that there are no substantial differences in the rationality of is “teach students in accordance with their aptitude.” This performance when considering variables such as gender, age, study may assist in efficiently guiding further policy reforms education or social status, which is partly due to a lack of in this direction. In addition, the findings also imply that the standardization and regularity in the methodologies used (e.g., subject’s perspective is an important factor in the decision- Gigerenzer and Murrey, 1987; Hertwig and Ortmann, 2001, making process. 2005). This experimental study examined the framing effect Finally, one potential issue is that the participants may have of Chinese college students with the aim of evaluating the only provided their estimation of how “certain figures” would effectiveness of HE reform from a different perspective. The respond to each scenario. The participants may lack adequate results show that the participants’ preferences were basically knowledge of these manipulated roles, which might undermine unbiased and at least partially driven by the high numeracy the results. In this regard, more preliminary work could be skills the participants had accumulated through higher education, conducted in an attempt to improve the research design in the which demonstrates the weakness of classical framing effects. future. Thus, this study offers new evidence on the relationship between education and behavior in China. Furthermore, this study sheds light on how such an analysis might help to elucidate evaluations AUTHOR CONTRIBUTIONS of the effects of China’s HE policy, under which more young people have access to higher education and receive better training WF designed and carried out experiments, as well as analyzed the in thinking and reasoning skills. China’s educational system results and wrote the manuscript. has clearly been expanding on a sustainable developmental path. When the student responses were sensitively manipulated, FUNDING however, dividing the sample by sex revealed a tendency for them to be influenced by framing effects. A significant gender I gratefully acknowledge financial supports from the Priority difference emerged in this context, which calls for more research Academic Program Development of Jiangsu Higher Education to deeply determine whether the framing effects are susceptible Institutions (PAPD); SRF for ROCS, SEM; Key Program of to individual differences and sheds light on improving teaching Humanity and Social Science foundation, State Education output. More consideration should therefore be given to teaching Ministry 16JJD840008; Nanjing University of Finance and styles and material as well as students’ aptitudes in light Economics Pre-research fund YYJ201413 and “Young Scholar” of the sex differences revealed in sensitivity to the adaptive Program. Harris, P. (2008). Designing and Reporting Experiments in Psychology, 3rd Edn. REFERENCES Maidenhead: Open University Press. Alwang, J., Larochelle, C., and Barrera, V. (2017). Farm decision making and Hasseldine, J., and Hite, P. A. (2003). Framing, gender and tax compliance. J. Econ. gender: results from a randomized experiment in Ecuador. World Dev. 92, Psychol. 24, 517–533. doi: 10.1016/S0167- 4870(02)00209- X 117–129. doi: 10.1016/j.worlddev.2016.11.015 Hertwig, R., and Ortmann, A. (2001). Experimental practices in economics: a Anderson, C. L., Reynolds, T. W., and Gugerty, M. K. (2017). Husband and methodological challenge for psychologists? Behav. Brain Sci 24, 383–403; wife perspectives on farm household decision-making authority and evidence discussion 403–451. doi: 10.1037/e683322011- 032 on intra-household accord in rural Tanzania. World Dev. 90, 169–183. Hertwig, R., and Ortmann, A. (2005). “The cognitive illusion controversy: a doi: 10.1016/j.worlddev.2016.09.005 methodological debate in disguise that matters to economists,” in Experimental Arcidiacono, D. (2011). Consumer rationality in a multidisciplinary perspective. Business Research, Vol. 3, eds R. Zwick and A. Rapaport (New York, NY: J. Socioecon. 40, 516–522. doi: 10.1016/j.socec.2011.01.003 Springer), 113–130. doi: 10.1007/0- 387- 24244- 9_5 Chambaere, K., Rietjens, J. A., Cohen, J., Pardon, K., Deschepper, R., Pasman, Huang, Y., and Wang, L. (2010). Sex differences in framing effects across task H. R. W., et al. (2013). Is educational attainment related to end-of-life decision- domain. Personal. Individ. Dif. 48, 649–653. doi: 10.1016/j.paid.2010.01.005 making? A large post-mortem survey in Belgium. BMC Public Health 13:1055. Kidd, A. M., Monz, C., D’Antonio, A., Manning, R. E., Reigner, N., Goonan, doi: 10.1186/1471- 2458- 13- 1055 K. A., et al. (2015). The effect of minimum impact education on visitor spatial Fan, J., and Zhang, C. (2015). A study of the rate of return to higher engineering behavior in parks and protected areas: an experimental investigation using GPS- education in China. Int. J. Educ. Dev. 42, 106–114. doi: 10.1016/j.ijedudev.2015. based tracking. J. Environ. Manage. 162, 53–62. doi: 10.1016/j.jenvman.2015. 04.010 07.007 Fan, W., Ma, Y. Y., and Wang, L. M. (2015). Do we need more public investment in Levin, I. P., Gaeth, G. J., Schreiber, J., and Lauriola, M. (2002). A new look higher education? Estimating the external returns to higher education in China. at framing effects: distribution of effect sizes, individual differences, and Asian Econ. Pap. 14, 88–104. doi: 10.1162/ASEP_a_00380 independence of types of effects. Organ. Behav. Hum. Decis. Process. 88, Gigerenzer, G., and Hug, K. (1992). Domain-specific reasoning: social contracts, 411–429. doi: 10.1006/obhd.2001.2983 cheating, and perspective change. Cognition 43, 127–171. doi: 10.1016/0010- Lusardi, A. (2008). Household Saving Behavior: The Role of Literacy, Information 0277(92)90060- U and Financial Education Programs. Hanover, PA: Dartmouth College. Gigerenzer, G., and Murrey, D. J. (1987). Cognition as Intuitive Statistics. Hillsdale, Maloney, M., and McCarthy, A. (2016). Understanding pension communications NJ: Erlbaum. at the organizational level: insights from bounded rationality theory & Frontiers in Psychology | www.frontiersin.org 6 May 2017 | Volume 8 | Article 744 fpsyg-08-00744 May 9, 2017 Time: 15:44 # 7 Fan Framing Effect in China implications for HRM. Hum. Res. Manage. Rev. 27, 338–352. doi: 10.1016/j. Tversky, A., and Kahneman, D. (1986). Rational choice and the framing of hrmr.2016.08.001 decisions. J. Bus. 59, S251–S278. doi: 10.1086/296365 Mendelberg, T., and Karpowitz, C. F. (2016). Women’s authority in political Wang, X. T. (1996a). Evolutionary hypotheses of risk-sensitive choice: age decision-making groups. Leadersh. Q. 27, 487–503. doi: 10.1016/j.leaqua.2015. differences and perspective change. Ethol. Sociobiol. 17, 1–15. doi: 10.1016/ 11.005 0162- 3095(95)00103- 4 Mitchell, O., and Lusardi, A. (2015). Financial Literacy and Economic Outcomes: Wang, X. T. (1996b). Framing effects: dynamics and task domains. Organ. Evidence and Policy Implication. Working Paper. Philadelphia, PA: University of Behav. Hum. Decis. Process. 68, 145–157. doi: 10.1006/obhd.1996. Pennsylvania. 0095 Nafría, J. M. D., Cendón, J. A., and Alonso, L. P. (2015). Building up eParticipatory Xu, Y., Xu, Z., and Zhang, J. (2000). The nursing education system in the People’s decision-making from the local to the global scale. Study case at the European Republic of China: evolution, structure and reform. Int. Nurs. Rev. 47, 207–217. higher education area. Comput. Hum. Behav. 47, 26–41. doi: 10.1016/j.cnb.2014. doi: 10.1046/j.1466- 7657.2000.00027 09.004 You, L. M., Wan, L. H., Liu, K., Zhu, X. W., Zheng, J., and Luo, Z. M. (2010). OECD (2009). OECD Reports on PISA Results. Available at: http://www.oecd- Education capacity of nursing schools and the needs to nurses in China. Nurs. ilibrary.org/education/pisa_19963777 Manage. 10, 46–50. OECD (2012). OECD Reports on PISA Results. Available at: http://www.oecd- Zhang, Y. L., and Miao, D. M. (2008). Social cues and framing effects in risky ilibrary.org/education/pisa_19963777 decisions among Chinese military students. Asian J. Soc. Psychol. 11, 241–246. Oehler, A., and Werner, C. (2008). Saving for retirement—a case for financial doi: 10.1111/j.1467- 839X.2008.00263.x education in Germany and UK? An economic perspective. J. Con. Policy 31, Zhang, Y. L., Xiao, L. J., Ma, Y., and Miao, D. M. (2008). Effect of framing on risky 253–283. doi: 10.1007/s10603- 008- 9074- 5 choice: one case study in China. Soc. Behav. Pers. 36, 651–658. doi: 10.2224/sbp. Peters, E., and Levin, I. P. (2008). Dissecting the risky-choice framing effect: 2008.36.5.651 numeracy as an individual-difference factor in weighting risky and riskless options. Judgm. Decis. Mak. 3, 435–448. doi: 10.2307/1914185 Conflict of Interest Statement: The author declares that the research was Petroman, C., Bidireac, I. C., Petroman, I., Sucan, M., Marin, D., Turc, B., et al. conducted in the absence of any commercial or financial relationships that could (2015). The impact of education on the behaviour of the consumer of animal be construed as a potential conflict of interest. origin food products. Procedia Soc. Behav. Sci. 190, 429–433. doi: 10.1016/j. sbspro.2015.05.021 Copyright © 2017 Fan. This is an open-access article distributed under the terms Saad, G., and Gill, T. (2014). The framing effect when evaluating prospective mates: of the Creative Commons Attribution License (CC BY). The use, distribution or an adaptationist perspective. Evol. Hum. Behav. 35, 184–192. doi: 10.1016/j. reproduction in other forums is permitted, provided the original author(s) or licensor evolhumbehav.2014.01.002 are credited and that the original publication in this journal is cited, in accordance Tversky, A., and Kahneman, D. (1981). The framing of decisions and the with accepted academic practice. No use, distribution or reproduction is permitted psychology of choice. Science 211, 453–458. doi: 10.1126/science.7455683 which does not comply with these terms. Frontiers in Psychology | www.frontiersin.org 7 May 2017 | Volume 8 | Article 744

Journal

Frontiers in PsychologyPubmed Central

Published: May 11, 2017

References