Chapter 11 Chapter 7 "But It's Not Real Life!" - "Artificiality" Criticism and Psychology

Having covered the principles of experimental logic in the previous two chapters, we can now consider some of the criticisms that psychology is often confronted with.For example, many people think that scientific experiments are worthless because they are artificial and different from "real life".We will explore this point of view in detail.Since psychological experiments are often subject to similar criticisms, understanding the irrationality of this criticism will help us better understand psychology. From the content of Chapter 6, we can already see clearly why this criticism is unreasonable.As noted in the previous chapter, the artificiality of scientific experimentation is not a shortcoming; in fact, it is what gives the scientific method its strange power to allow us to explain the world.Contrary to what is often believed, the artificiality of scientific experiments is not an accidental oversight but a deliberate act by scientists.The reason why scientists specifically set some non-naturally occurring conditions is that only in this way can they distinguish the many related variables that determine the occurrence of events.Sometimes the necessary conditions already exist in the state of nature, as in the case of Snow and cholera.But that doesn't happen very often.Scientists must manipulate events in novel and sometimes bizarre ways, as in the case of Goldberg and pellagra.Many times, these operations cannot be done in the natural environment, so scientists must transfer the phenomenon to be studied to the laboratory for more precise control.For example, in the early research on "gravity and motion", some special objects were used, the purpose of which was to create some special conditions in order to observe the motion of objects.Therefore, in order to analyze a phenomenon, it is often necessary to create unnatural extreme conditions.

In fact, some phenomena would be impossible to discover if scientists were completely confined to observe under "natural" conditions.Physicists exploring the nature of matter build giant accelerators to induce collisions between elementary particles.Some byproducts of the collisions are new particles that exist for less than a billionth of a second.However, the properties of these new particles help explain the theory of atomic structure.Many new particles generally do not exist in the world, and even if they exist, we have no chance to observe them under natural conditions.As a result, few question the way physicists work.In order to gain a deeper understanding of the universe, it is reasonable to adopt even some uncommon and even weird methods.But for some reason, methods that seem reasonable to physicists are often considered irrational to use by psychologists.

Scientists of all kinds have been misunderstood by the public's lack of awareness of the importance of creating special conditions, and psychologists have been the greatest victims of this misunderstanding.Many psychologists have heard the sigh, "Too bad it's not real life," after showing experimental evidence about a behavior to a layperson.Further discussion of this critique often brings us to popular beliefs, such as that knowledge can only be gained by studying natural conditions, that laboratory research in psychology is weird and not good enough to be a science. Many of the techniques psychologists use are weird to the public, and many people don't know that these techniques are not unique to the field of psychology, but psychologists apply these scientific methods to the study of human behavior.In fact, in surveys of every discipline, one sees similarly bizarre ways in which people acquire knowledge about the world.Psychologists have been hit on both sides.Ignorance of the scientific method led many to believe that psychology could never be a science, and psychologists, like all other scientists, created the special conditions necessary to give stronger and more precise explanations of psychological phenomena. defamation.

Imprisonment to real-life conditions prevents us from discovering many new things.For example, biofeedback techniques are now widely used in various fields, such as for the control of recurrent migraine and tension headaches, the treatment of high blood pressure, and relaxation training (deCharms et al., 2005; Maizels, 2005; Miller, 1985) .Studies have shown that humans can learn to control these processes to a certain degree if they can monitor ongoing physiological processes in the body through visual or auditory feedback.This research has contributed to the development of the aforementioned biofeedback technology.Of course, because humans do not have the ability to monitor their own physiological functions through external feedback, it will be difficult for people to find that humans have the ability to control their own physiological processes if it is not under special laboratory conditions.Observation under natural conditions would never have discovered this.

Sometimes, however, complaints like "This is not real life" stem from another misunderstanding of the purpose of psychological experimental research, and the reasons for this misunderstanding are very easy to understand.The media campaign has familiarized many people with survey research, especially opinion polls in elections.There is now a growing understanding of some important features of electoral voting.Specifically, in order to ensure the accuracy of polls, the media pays more attention to concepts such as random sampling and sample representativeness.This focus has led many to mistakenly believe that random sampling and representativeness are necessary for all investigative research in psychology.Because psychological research rarely uses random samples of subjects, many psychological research results will be discredited if it is based on random sampling standards that laymen believe. Those who criticize psychological research cannot reflect real life and are therefore invalid. Arguments are also strengthened.

But it is easy to understand the absurdity of this idea by thinking of other sciences.Chemists have never tried to take random samples of chemical compounds, and biologists have never experimented with random samples of cells or tissues.Mice and monkeys used in medical research are also not fully representative of their species.And these studies were conducted in laboratories that were completely different from the natural environment in which these animals lived.In fact, these conditions are often quite unique.However, the results of all these studies can help us understand human biology.The same is true for most psychological research.Not every psychological survey requires the use of random samples.Therefore, the important point we need to emphasize here is that random sampling and random assignment (see Chapter 6) are not the same thing.

Both random assignment and random sampling have "random" in them, so many people think they mean the same thing.In fact, they are very different concepts, the only similarity is that they both use randomly generated numbers.Its purpose, however, is quite different. Random sampling involves how subjects are selected for research.As mentioned earlier, not all research requires random sampling, but when it is necessary (for example, in survey research, market research, or polls during elections), we need a way to draw random samples from the population. To draw a sample, this method ensures that every member of the population has an equal opportunity to be selected as a sample, and the selected sample becomes the subject in the subsequent investigation and research.It is very important to note that such a randomly sampled survey study may be both a correlational study and a true experiment.Only by using random allocation can it become a real experiment.

Random assignment is a necessary condition for a true experiment.The experimenter divides the subjects into the experimental group and the control group. When each subject has the same chance of being assigned to the experimental group as to the control group, random assignment is achieved.To achieve this, randomization such as flipping a coin (and more often a special randomized number table) is often used - because it does not have any bias in grouping subjects. Random assignment and random sampling are not the same thing, and the best way to keep this in mind is to figure out four combinations: a nonrandom sample of nonrandom assignment, a nonrandom sample of random assignment, a random sample of nonrandom assignment, and a random assignment of random sample.Most psychology experiments do not use random samples because there is no need for them.As we'll see in the next chapter, research can test theories, and all we need is a handy sample.A study is a true experiment if random assignment is used, and a related investigation if it is not.Many studies that use random sampling do not use random assignment, and that's because they are survey studies looking for associations—that is, they are related survey studies.However, some studies that use both random sampling and random allocation must be true experiments.

University of Virginia psychologist Douglas Mook articulates the different types of predictions that different types of research require.The goal of much applied research is to relate findings directly to specific situations in life.For applied research, the prediction must be: there is a "one-to-one" relationship between research and real life, which is what Muck calls "analogy", and the results of applied research must be directly applicable.Polls in election polls are an example of applied research.The purpose of the research is to predict a specific behavior in a specific situation, in this case, the voting results on Election Day.Since the research results are to be directly applied to reality, the randomness of the sample and the representativeness of the situation are very important.

However, it would be a mistake to regard applied psychology research as typical psychology research.Most research in psychology (or other disciplines, for that matter) has a purpose other than application.The level of prediction in most research is from theory to prediction for a specific research situation.The results of most research can only be applied indirectly through the modification of theories, which are applied together with other scientific laws to some practical problems (Nickerson, 1999).In short, most theoretical research pursues the theoretical verification of psychological processes, rather than extending the research results to a special situation in reality.

Research whose main purpose is theory verification is often referred to as "basic research".The purpose of applied research is to apply data directly to real life, but basic research focuses on theoretical verification.However, it is likely to be erroneous to distinguish between basic and applied research solely on the basis of whether a certain research has practical applications, since this distinction often fades over time.The results of applied research are quickly applied.But nothing is more practical than a general, accurate theory.Although the original intention of many scientists to carry out theoretical or empirical research is not to solve specific practical problems, the scientific theories or research results they develop eventually solve many problems in the real world.Such examples abound in the history of science. Robert Crease & Nicholas Samios (1991), working at Brookhaven National Laboratory, a leading physics research center, reported in a paper Some examples of basic research having applied value are presented in .Take, for example, the story of Wilhelm Roentgen, who "discovered a wonderful thing, that when he placed a phosphorscreen close to his apparatus, it unexpectedly glowed; A new phenomenon, which he called X-rays. Three months later, X-rays were already being used to detect broken bones" (p.82).Or Howard Florey, who, with a colleague, was working on antimicrobial mechanisms. "Penicillin is one of the microbes they studied, which was discovered by accident a few years ago, but its antibacterial function has not been discovered. Many drugs today are discovered in some academic research with no special interest in application. It was developed on the basis of "(p.82) Seymour Kety (1974) also talks about how ciprofloxacin, a specific drug for schizophrenia, was discovered in seemingly unrelated scientific studies.Caddy emphasizes that almost all discoveries that apply to schizophrenia treatments have nothing to do with schizophrenia!History has demonstrated time and again that attempts to control the direction of scientific development (by having scientists solve specific practical problems) can only impede progress rather than facilitate it.Ironically, rushing scientists to solve practical problems without making them think about "other things" (basic research) has proven to be the most impractical and short-sighted. The road to practical application is very winding and unpredictable.To study arthritis, a team of researchers at the Medical Research Center at Southwestern University in Texas tried to genetically breed mice with arthritis.Unexpectedly, these mice also developed enteritis (Fackelman, 1996).This research group "accidentally created mice with ulcerative colitis, thereby providing an animal model for scientists to study the human disease" (Fackelman, 1996, p. 302).Whether or not these scientists made progress on arthritis (the problem they were supposed to study), it now appears that they have made a huge contribution to the treatment of ulcerative colitis. In short, we must be aware that while some research is designed to directly predict a particular situation, most scientific research is fundamental research used to test a theory.How to apply the research results to real life?Researchers engaged in applied research and those engaged in basic research have different answers to this question.The former will answer like this: "Direct application, as long as the experimental situation is quite similar to the situation to be applied in the future." Therefore, the random sampling of the subjects and the representativeness of the experimental situation will affect the application of the results.However, the researcher who tests the theory assumes that the results of the study will not have direct application to real life, nor is the purpose of carrying out theoretical research to apply the results to specific environmental conditions.Thus, this type of scientist is not concerned with how similar the subjects in the study are to other groups, or whether the experimental situation reflects some real-life situation.So, does that mean these findings have no real-world implications?no.These findings are not directly applicable to a particular situation, but to theory.This theory, perhaps one day in the future, can be combined with other scientific laws to jointly solve a special problem. In some fields of psychology, this kind of indirect application of theory to real life research is very common.For example, when cell phones were first introduced many years ago, many cognitive psychologists immediately began to worry about safety—what would happen to people talking on their cell phones while driving.Psychologists immediately predicted that cell phone use might lead to an increase in traffic accidents—not just because one hand was taken off the steering wheel to answer a call, but also because they worried that answering a call would distract drivers.It's important to realize that psychologists raised these concerns long before actual experimental studies using mobile phones (see Strayer & Johnston, 2001).Psychologists have theorized to predict cell phone accidents, and in this case attention-limited processing theory has been around for decades (eg, Broadbent, 1958; Kahneman, 1973).This theory of information processing has been established through a large number of experimental verifications (hundreds of laboratory studies), and using a mobile phone while driving provides an opportunity to use this theory to predict its possible harm.This is also the case, and the later research using mobile phones confirmed the predictions of attention theory in psychology: the use of mobile phones is indeed a cause of traffic accidents (Insurance Institute for Highway Safety, 2005; Mc Evoy et al., 2005; Redelmeier & Tibshirani, 1997, 2001). Douglas Mock (1983) illustrates the idea of ​​experiment-testing theory in psychology and the nature of its indirect application as an example. In the 1930s, Selig Hecht published a series of studies on visual sensitivity in the Handbook of General Experimental Psychology (Murchison, 1934), speaking of dark adaptation The phenomenon.You may have experienced temporary "blindness," such as when you walk into a dark movie theater.However, after you've sat in position for a while, you should notice that chairs, people, and other objects slowly become visible.If you continue to pay attention to this phenomenon, you will find that this process of increasing visual acuity lasts for several minutes. This phenomenon is called dark adaptation, and it goes through two stages: first, there is a small, rapid increase in visual acuity when entering a dark room, followed by a slow, large increase.Hecht related the two-part ascending curve to the two types of photoreceptor cells in the retina.Cone cells densely packed in the center of the fovea (the part of the retina that focuses light) are very sensitive to red light.The rods, distributed around the fovea, are less dense and less sensitive to red light.From these facts, Hecht theorized that the initial phase of dark adaptation (a small, rapid increase in visual acuity) resulted from the adaptation of the cones, and that the second phase (a large increase in visual acuity over a longer period High) stems from rod adaptation. Mook (1983) reminds us to consider Hecht's experimental setting as completely unnatural. (Not randomly sampled) Participants responded in a dark room, answering "Yes, I can see it" or "No, I can't see it," depending on whether they detected a faint red flash.Normally, we don't respond "yes" or "no" to faint red lights in our daily lives.But since Hecht isn't thinking about extending his findings to individuals who responded "yes" or "no" to red light in a dark room, it doesn't matter whether this actually happened in real life. .What Hecht is concerned about is how to verify the corresponding theory based on the facts established in the laboratory, so as to explain some basic processes unique to the visual system, such as dark adaptation.He was not concerned with whether his experimental situation corresponded to real-world conditions, but with whether he could sufficiently isolate the particular visual process he wanted to study. Hecht's findings are general not because the nature of his experimental situation was artificial or natural, but because they allow a theory of fundamental visual processes to be developed that can be compared with Many visual phenomena are associated.His research revealed the functional relationship between the various parts of the human visual system precisely because his research situations were precisely controlled by humans.If the theoretical model is correct, it should be applicable in a wide variety of contexts and can be used to explain many behavioral phenomena, even in very different contexts than the one in which the theory was discovered.In other words, Hecht's research results have an indirect application value through their impact on theory.For example, Hecht's findings led to treatments for night blindness and improved X-ray recognition problems (Leibowitz, 1996; Mook, 1982).More notably, British aviators during the Blitzkrieg during WWII while awaiting the nighttime raids of Hitler's bombers wore red flying glasses (because the rods are not sensitive enough to red light to maintain dark adaptation; see Mook, 1982 ).From judging little red dots in a laboratory to discerning the movement of dangerous objects over London, the gap was bridged by theory, not by turning Hecht's laboratory into a fighter jet. Once we understand that the purpose of most research is to develop theories rather than predict events in specific circumstances, and that the results of most research are indirectly applied through theory rather than direct application in specific environmental conditions, then we will naturally understand Ask: How many theories in psychology can be applied in reality.That is, has the universality of psychological theories been verified? At this point, we must admit that the past record has been spotty.But it must also be clearly realized that this is closely related to the diversity of psychology.Research in some areas has indeed made little progress in terms of application, while others have achieved remarkable results, and many principles with explanatory power and predictive power have been derived through experiments (see Buss, 2003; Byrnes, 2001; Davidson & Sternberg, 2003; Gazzaniga, 2004; Hilton, 2003; Sunstein, 2002; Wilson & Keil, 1999; Zimbardo, 2004).For example, in applied fields such as counseling, education, clinical practice, and psychotherapy, the results of theoretically oriented basic research have also gained a large number of practical applications (Davidow & Levinson, 1993; Garb & Schramke, 1996; Lilienfeld, Lynn, & Lohr, 2003; Lilienfeld , Wood, & Garb, 2000; Mahrer, 2000; Siegert & Ward, 2002). Think of the principles of classical conditioning and operant conditioning.These principles, and the laws they describe in detail, were developed almost entirely from experiments with non-human subjects, such as pigeons and rats, in highly artificial laboratory environments.However, these principles have been successfully used to solve a wide variety of human problems, including the treatment of children with autism, the teaching of large amounts of factual material, the treatment of alcoholism and obesity, the management of patients in mental institutions, and the treatment of phobias etc.This is only a small part. The principles underlying these applications can be refined precisely because, under laboratory conditions, researchers are able to refine the relationship between environmental stimuli and behavior with precision that cannot be done under natural conditions. Because in natural situations, relationships between many behaviors may be at work simultaneously.As for the use of non-human subjects, this is because in many cases theories and patterns derived from animal responses provide us with data that closely resemble human behavior (Vazire & Gosling, 2003).Human studies have found that human behavioral patterns are very similar to those derived from animal behavior.These findings should not surprise us when every medical advance in the treatment of human disease has been derived from data from animal studies.For example, animal research has led to advances in many fields, including behavioral medicine, stress relief, psychotherapy, rehabilitation of the injured or disabled, the effects of aging on memory, ways to help people overcome neuromuscular disorders, and the effects of drugs on fetal development. effects of drug abuse, memory loss, traffic safety, and treatment of chronic pain (Domjan & Purdy, 1995; Gosling, 2001; Kalat, 1995; Miller, 1985b; Zimbardo, 2004).More recently, research on dogs has provided a substantial impetus to understanding the fundamentals of anxiety disorders in humans (Groopman, 1999). In fact, the "it's not real life" criticism has been wrongly used to discredit animal research—often for political reasons.For example, politicians who work for heavily polluting companies consistently deny the validity of carcinogen risk assessment reports on the grounds that they are based on animal studies and cannot be applied to human risk assessments.However, a group of scientists in a 1988 study of 23 carcinogens (benzene, asbestos, etc.) found that the death rates calculated from animal studies were very similar to those calculated from human epidemiological studies ( Rnkel, 1996). Psychologists have made impressive progress in the study of perceptual processes, and the laws and theories derived from them have been used to solve problems as diverse as radar monitoring systems, street lighting, and the design of aircraft cockpits (Nickerson , 1992; Swets, Dawes, & Monahan, 2000; Wickens, 1992).Much new knowledge has been accumulated about the effects of aging on cognition (Salthouse, 2004), and this new knowledge may directly help us design training programs to help those with cognitive loss restore their abilities (Dixon & Backman, 1995). Psychological research on judgment and decision-making has been applied to decision-making in areas such as medicine, education, and economics (Gigerenzer, 2002; Gilovich, Griffin, & Kahneman, 2002; Hilton, 2003; Kahneman, 2003; Kahneman & Tverslgr, 2000; Swets, Dawes , & Monahan, 2000; Tetlock, 2005; Zweig, 2001).An exciting development is the increased involvement of cognitive psychologists in the legal system, which opens up numerous opportunities for the application of cognitive theory in information gathering, evidence evaluation, and decision making (Kassin , Tubb, Hosch, & Memorx, 2001; Koehler, 1993; Kuhn, Weinstock, & Flaton, 1994; Redding, 1998; Wrightsman, 1999; Zimbardo, 2004).Moreover, from the mid-1980s, the theory and practice of reading instruction began to be influenced by cognitive psychology (Rayner, Foorman, Perfetti, Pesetsky, & Seidenberg, 2001, 2002; Snowling & Hulme, 2005; Stanovich, 2000; Veilutino, Fletcher, Snowling, & Scanlon, 2004).There are also questions about the accuracy of children’s testimony in legal proceedings (Ceci & Hembrooke, 1998) and the accuracy of “recovered” memories of abused children (Brainerd & Reyna, 2005; Bremner, Shobe, & Kihlstrom, 200; Clancy, Schacter, Mc Nally, & Pitman, 2000; Gardner, 2006; Loftus, 1997; Lynn, Loftus, Lilienfeld, & Lock, 2003; Mc Nally, 2003; Pezdek & Banks, 1996; Spanos, 1996) and other issues of public debate, Psychologists have provided important scientific evidence.The American Psychological Association maintains a website: www.psychologymatters.org.You can see more practical applications of psychology knowledge on this website. Many people question the representativeness of psychological research results. They pay too much attention to the subjects of the research and do not care about the details of the experimental design.However, in many areas of psychology, research results are more influenced by the latter than by the former, such as the study of fundamental processes of perception.In terms of basic information processing, the basic structure of the brain, and the characteristics of the visual system, people in Montana and people in Florida are very similar.Also, these characteristics had no bearing on whether the participants' parents were tinkerers, tailors, or professors. All disciplines assume that certain factors have no influence on the final outcome.Biologists are generally not concerned that small differences in the thickness of a petri dish will significantly affect the bacteria inside.Of course, these differences may have an effect—every scientific hypothesis is not absolute—but biologists must focus on the variables they think are more likely to have an effect.Similarly, Hechter assumed that implicit adaptation had nothing to do with a person's religion, so he did not ask whether the subjects were Lutheran or Roman Catholic. We face a problem, sometimes called the "sophomore problem"; namely, that because sophomores are subjects in a large number of psychological studies, the generalizability of the results of these studies is questioned.Psychologists care about this question because it is a problem in some fields of study.Still, we want to put it in perspective and know that psychologists have several legitimate defenses for this criticism.Three points are listed below: 1. This criticism does not mean that the research results are invalid, but that more research is needed to prove the generalizability of the theory.Since we previously collected data on sophomores, even having contrary data from other populations necessitated adjustments to the theory that would only make the theory more precise, not negate it entirely.Even in the more extreme cases where repeated experiments do not yield the same results, we can only say that the theory based on the sophomore data is not comprehensive enough to say that the theory is necessarily wrong. 2. In many fields of psychology, the sophomore problem does not constitute a problem, because the psychological process studied is a very basic process (such as the visual system), and almost no one believes that the basic structure of the visual system is the same as that of the sample. related to the characteristics of the population distribution. 3. Many of the findings have been replicated, leading us to believe that the results generalize to a large extent to different geographic distributions and, to a lesser extent, to individuals with different socioeconomic factors, family variables, and early among people with educational experience. Fifty years ago, the sample of college students probably came from elite groups, but now it is completely different. Now the family background of college students can represent groups from all walks of life. However, it would also be unwise not to acknowledge that the sophomore problem is indeed a problem in some areas of psychological research (Peterson, 2001).Nonetheless, psychologists are doing their best to correct the problem.For example, developmental psychologists are almost all concerned with this question.Every year, hundreds of researchers in this field re-verify many theories and findings obtained with college students on other subjects of different ages. Using different age groups as subjects does not always reproduce the results obtained with university students.If that were the case, developmental psychology would be boring.But a whole host of psychologists have worked to establish an age factor in psychological theory to demonstrate its importance, and research in this area has also ensured that grand theories of psychology are not based solely on gleaned from college students. based on limited data. Many studies of developmental processes have been conducted with North American children. To assess the generalizability of these findings, developmental psychologists also conduct cross-cultural studies.For example, Stevenson et al. (1985) had Chinese, Japanese, and American children perform a number of cognitive tasks and concluded that "children from all three cultures performed Cognitive abilities are very similar” (p.727>. Cross-ethnic and cross-cultural comparative studies of other cognitive abilities have also confirmed the same results (Demetriouetal., 2005; Mc Bride-Chang & Kail, 2002). For other psychological characteristics, there are many examples in cross-cultural studies that show similar trends (eg, Day & Rounds, 1998; Rozin, Lowery, Imada, 1999). However, there are also many cross-cultural studies that do not show similar trends with American sophomores. (e.g., Nisbett, 2003). But when these differences are present, these studies still provide some important information about how these theories and results can vary across cultures and contexts (Chang, 1996; Menrich et al. , 2004; Medin & Atran, 2004; Nishett, 2003; Yates, Lee, & Shinotswka, 1996). As mentioned earlier, research findings in cognitive psychology have been replicated.Many fundamental laws of information processing have been validated in many laboratories around the world.People may not know that if a psychologist at the University of Michigan achieves a major research result, similar experiments will soon be performed at Stanford University, University of Minnesota, Ohio State University, Cambridge University, Yale University, University of Toronto, etc. University conducts.Through this test, we will quickly know whether this result is due to the uniqueness of the Michigan subjects or the special experimental environment. Educational psychologists have also asked the sophomore question.For example, educational psychologists, in conjunction with developmental psychologists and other educational researchers, have constructed measures of basic cognitive skills that predict academic achievement with some accuracy, such as reading acquisition rates.The predictive accuracy of these content was not affected by socioeconomic status and race, nor by geographic distribution and school district. Sophomore questions and critiques of representation are largely directed at social psychology, which is often studied in laboratory settings using undergraduate subjects and attempts to establish social interactions, group behavior, and information in real social situations processing and other theories (Kimda, 1999).Even in this area of ​​psychology, however, there is evidence that lab-derived results and theories do, in fact, predict the behavior of different types of individuals in a variety of situations. For example, a few years ago, Leonard Berkowitz, a psychologist at the University of Wisconsin, demonstrated what is known as the "weapon effect"—the fact that the presence of a weapon makes someone It is easier to react aggressively.This discovery originated in the laboratory and is a typical example of an unrepresentative situation.Since this result is an induced product of an artificial situation, it has often been strongly criticized as misleading.But the fact is that the results are the same in all experimental conditions, the same results are obtained in different ways of measuring aggression, the same results are obtained in Europe and the United States, the same results are obtained in studies of children and adults, in the laboratory The same results were obtained in field studies other than those in which subjects were unaware that they were participating in the experiment (Berkowitz & Dormerstrein, 1982; Turner, Simons, Berkowitz, & Frodi, 1977).The researchers even extracted the cognitive mechanisms behind the weapons effects.在语义记忆中，它是一个自动启动的过程（见Anderson, Benjamin, & Bartholow, 1998）。这个效应是如此之强，以至于安德森等人（Anderson et al., 1998）把文章题目定为“是手枪自己扣动扳机的吗？” 认知、社会和临床心理学家也研究了人类的各种决策行为。这个研究领域里大部分原始的研究都是在实验室里完成的，使用大学生作被试，而且采用高度人为化的任务。然而，从这些研究中得出的决策行为原则在很多非实验室环境中都得到了重现，包括银行家对股票价格的判断、赌场赌博、精神病医生对病人行为的预测、经济市场预测、军事情报分析、全美橄榄球联赛的下注、工程师对修理时间的估计、房地产经纪人对房价的估计、商务决策以及医生的诊断——这些原则现在也应用于个人理财咨询的实践领域（Bazerman, 1999, 2001; Belsky & Gilovich, 1999; Goldstein & Hogarth, 1997; Hammond, 1996; Tassoni, 1996; Zweig, 2001, 2002）。 伯尔尼鲍姆（Birnbaum, 1999）用互联网来解决心理学中的大二学生问题。他在实验室里通过互联网招募了一批参与者，并进行了一系列有关决策问题的实验。实验室中得到的结果全部在互联网样本中得以重现，而后者的取样范围要比前者广泛得多——包含来自44个国家的1224名参与者（见Jaffe, 2005; Mc Graw, Tew, & Williams, 2000）。高斯林等人（Gosling et al., 2004）研究了大量互联网参与者的样本（361703人），并将之与发表过的510个传统样本的参与者比较，发现互联网上的被试在性别、社会经济地位、地区和年龄方面有着更广泛的分布。重要的是，他们发现，心理学众多研究领域的研究结果，例如人格理论，用互联网实验和传统方法的研究所得出的结果非常相似。 这些例子说明，心理学研究结果的稳定性和普遍性程度经常被低估了（Cheng, 2001; Gage, 1996; Rosenthal, 1990）。安德森、林德西和布什曼（Anderson, Lindsay, & Bushman, 1999）对实验室研究和现场研究的效果进行了系统检验。从验证攻击性、领导力、抑郁症等多种主题的研究中，安德森等人（1999）发现了高度的聚合性——几乎所有不同的研究环境总是产生相似的结果。 当然，并不是所有的心理学研究成果都能重现（见Carroll & Nelson, 1993）。相反，结果无法重现的实验经常出现，而它们往往比结果可以重复的实验更具指导意义。但是，在认知心理学中，重复实验的失败几乎很少是被试的独特性造成的。相反，大部分是因为实验刺激和方法的细微差异。通过仔细地分析要产生一个现象究竟需要哪些实验条件，科学家们对现象有了更精确的理解，这为建立一个更精确的理论奠定了基础。 但是，如果实验结果没有被重现，那么心理学的研究成果如何应用？如果科学家们没有在所有的细节上达成一致，知识和理论并不完全站得住脚，那么如何证明这些结果的应用是合理的呢？这种对心理学发现的担心是很常见的，因为人们没有意识到，在其他科学中，结果和理论经常在完全确立之前就开始应用了。当然，第2章中已经清楚地阐述过，所有的科学理论都有可能被修订。如果我们在应用科学研究结果之前必须确定知识是完全正确的，那么应用就不会发生了。所有领域的应用型科学家尽最大努力使用最准确的信息，同时也会意识到这些信息有可能是错误的。 本章提到了几个焦点问题，此外有一点很重要，就是我们应该清楚什么是我们说过的，什么是我们没说的。我们证明了对心理学研究的频繁抱怨源自一个基本的误解，不是针对心理学，而且针对涉及所有科学的一个基本原则。我们看到人们质疑心理学家为什么不在所有研究中都使用随机样本，并且解释了这种担心是多余的。最后，我们看到，大二学生问题本来是一种合理的关注，但它有时被夸大了，尤其是当人们对心理学研究的广泛性和多样性不太熟悉时（见第1章）。 尽管如此，心理学家应当始终注意他们的实验结论不要太过依赖于某一种方法或某一特殊被试群体。这一点将在下章讨论。事实上，心理学的一些领域确实被大二学生问题折磨得够呛（Jaflfe, 2005; Peterson, 2001; Wintre, North, & Sugar, 2001）。作为大二学生问题的一剂良药，跨文化心理学仍然是一个亟待发展的领域。然而，研究型心理学家对于自我批评的高度重视，给了我们一个对此持乐观态度的理由（见第12章; Anderson & Gunderson, 1991; Henriques, 2003, 2004，2005; Jaffe, 2005; Kimble, 1999; Machado, Lourenco, & Silva, 2000; Proctor & Capaldi, 2001）。事实上，许多心理学家因为将“批判”作为其本职工作而享誉学界（Leary, 2001; Robinson, 2001）。每年的各类科学杂志上，都会有文章提醒心理学者注意其方法上的漏洞，或是指出大二学生问题。后者在心理学中是一个受到广泛关注的问题，没有心理学者尚未意识到这一点。因此，尽管我们不应忽视这一问题，同时也应正确看待它。 一些心理学研究属于应用型研究，它们的目标是把研究结果直接应用于特定情境。在这样的应用研究中，研究的目的是要将结果直接推广到自然情境中，样本的随机化和条件的代表性就显得尤为重要，因为研究结果将会直接得到应用。然而，大多数心理学研究不属于这种类型，而是属于基础研究，用以验证有关行为潜在机制的理论。在大部分基础研究中，研究结果通过理论上的修正得到间接应用，从理论产生到应用于某些实践性问题需要一段时间。在这种类型的基础研究中，被试的随机取样和情境的代表性不是关键问题，因为这类研究的重点在于验证理论的普遍性。实际上，在用于验证理论的基础研究中，人为的环境条件是有意创设的，因为（正如前一章所描述的）这有助于把研究的关键变量从所要控制的无关变量中分离出来。因此，心理学实验“不像是真实的生活”这个事实其实是一种优势而非缺点。