CrabHao火雲豪

At a minimum, one can confidently say that the relationship between discrepancy and persuasive effectiveness is not simple. Some investigations have found that-at least under some conditions-greater discrepancies are associated with greater effectiveness (i.e., a positive relationship between discrepancy and attitude change; e.g., A. R. Cohen, 1959; Hovland & Pritzker, 1957). But other studies have reported that, at least in some circumstances, with increasing discrepancy, persuasive effectiveness is reduced (a negative rela­tionship; e.g., A. R. Cohen, 1959; Hovland, Harvey, & Sherif, 1957).

至少可以確定的是差異性與說服效果之間的關係並不單純，有些研究至少在某一些情況下顯示，差異性愈大效果愈好，但也有些研究至少在某一些情況下顯示，當差異性逐漸增加時，效果則逐漸減少。

The most plausible general image of the relationship of discrepancy and effectiveness is that of an inverted-U-shaped curve, such that relatively little change is obtained with extremely small or extremely large discrepancies, and maximum effectiveness is to be found with moderate levels of discrepancy (for a general discussion of this view, see Whittaker, 1967; for findings of such curvilinearity-at least under some conditions-see, e.g., Aronson, Turner, & Carlsmith, 1963; Freedman, 1964; Sakaki, 1980; M. J. Smith, 1978; Whittaker, 1963, 1965). That is, with increasing discrepancy, there is increasing attitude change-up to a point (the peak of the curve); beyond that point, increases in discrepancy are associated with decreases in attitude change. This general curvilinear model can accommodate otherwise inconsistent findings of positive and negative relationships. For instance, if a given experiment has only rela­tively small levels of discrepancy, the results might well seem to suggest that increasing discrepancy leads to increasing attitude change (because there wasn't a sufficiently large discrepancy used in the research). Similarly, an experiment using only relatively large discrepancies might yield results indicat­ing that increases in discrepancy lead to lessened effectiveness.

對於差異性與效果間的關係，最佳的類比就是倒U字型，顯示當差異性最大化或最小化時，效果不顯著，當差異性最適宜時，效果最顯著。那就是當差異性逐漸增加時，態度的改變就增加到曲線的頂點；超過了這個點以後，差異性逐漸增加時，態度的改變卻相繼減少。這種曲線模式可以適用於正、反兩種不一致關係；例如，一個運用差異性小的實驗中，結果可能建議提高差異性以增加成效(因為在實驗中，缺乏夠大的差異性)；相同的，一個運用差異性大的實驗中，可能導引出差異性大卻減少成效的結果。

But even this curvilinear conception of the discrepancy-effectiveness rela­tionship requires some complications. It is not likely that there is just one curve that describes the relationship; rather, it's likely that there is a family of curves all having the same general inverted-U shape, but with the point of inflection in the curve (the peak, the point at which the curve turns back down) coming at different discrepancies. In some circumstances, that is to say, the point of maximum effectiveness may come at a relatively small discrepancy, but in another situation the point of maximum effectiveness may be at some larger discrepancy. From this point of view, then, it is not enough to know that the general shape of the discrepancy-effectiveness relationship is that of an inverted U. What is crucial is knowing what factors influence the location of the point of inflection in that curve - knowing the circumstances under which the point of maximum effectiveness occurs at relatively small discrepancies, and those under which it occurs at relatively large ones.

但是即使差異性與效果間關係的曲線概念有些複雜，並非僅有一種曲線可以解釋差異性與效果之間的關係；而是來自不同差異性的一個全都擁有轉折點(一個曲線迴轉的頂點)的倒U字型的曲線家族。在某些情況下，最大效果的點也許相對於最小的差異性，但是在另一個環境之下，最大效果的點也許相對於最大的差異。從這個觀點而言，卻無法充分了解差異性與效果之間關係的倒U字型曲線。重要的是瞭解何種因素將影響轉折點位置，瞭解相對於小的差異性，哪一個曲線的點可以發揮最大的功效，以及相對於大的差異性，哪一個曲線的點可以發揮最大的功效。

There look to be (at least) two main factors influencing the point of inflection in the discrepancy-effectiveness curve. One is communicator credibility. The peak of the curve appears to occur at smaller discrepancies for low-credibility communicators than it does for high-credibility communicators (see, e.g., Aronson et aI., 1963; Bergin, 1962; Bochner & Insko, 1966). That is, the optimal level of discrepancy is likely to be somewhat greater for a high-credi­bility communicator than for a low-credibility communicator; high-credibility sources can safely advocate somewhat more discrepant positions than can low-credibility sources.

最少有兩個主要的因素會影響差異性與效果的倒U字型曲線的轉折點，一個是溝通者的信賴度，對於低信賴度的溝通者時，曲線的頂端出現於較小的差異性，比起低信賴度的溝通者，高信賴度的溝通者則適合搭配較大的差異性。

The second is the receiver's degree of involvement with the issue. The research evidence suggests that for relatively high-involvement issues the peak of the curve occurs at lower levels of discrepancy, whereas on relatively uninvolving issues, the curve peaks at some larger discrepancy (see, e.g., Freedman, 1964; Sakaki, 1980). In a sense, then, as the issue becomes more and more personally relevant to the receiver, discrepancy becomes less and less tolerable. This effect of involvement was anticipated by social judgment theory (see Chapter 2). As receivers become more and more involved in an issue, their latitudes of rejection become larger, thus making the range of "objectionable" discrepant positions larger. For low-involvement receivers, a persuader may-be able to advocate a quite discrepant view without encountering the (small) latitude of rejection; but for high-involvement receivers, a highly discrepant view is likely to fall in the (large) latitude of rejection.

第二個則是受眾涉入相關議題的程度，研究顯示高度涉入的議題，曲線的高點在差異性較小時，相對的當議題的涉入度低時，曲線的高點在差異性較大時。就某種意義而言，當議題愈來愈與個人相關，差異性就愈來愈不容易被接受，涉入程度的效果在社會判斷理論中會顯現，當受眾對於相關議題的涉入程度愈來愈深的時候，反對的機率就相對變大，使得可反對的差異性位置變大；對於低涉入程度的受眾，說服者可能倡議差異性很大的議題，卻不會面臨反抗的態度；但是，對於一個高度涉入的受眾，極大差異性的意見極有可能面臨反對的聲浪。

FEAR APPEALS

The use of fear as a persuasive technique is an altogether common one. "If you don't do what I recommend," the communicator suggests, "then these terrible, fearful consequences will befall you." So, for example, high school driver education programs may show films depicting gruesome traffic accidents in an effort to discourage dangerous driving practices (such as drinking and driving); stop-smoking messages may display the horrors of lung cancer; dental hygiene messages may depict the ravages of gum disease; and so on.

The effectiveness of such "fear appeal" messages has been studied exten­sively. The central question that researchers have addressed concerns just how strong the fear appeal should be: Are stronger fear appeals more effective than weaker ones, or vice versa, or is there perhaps no general difference between them?

三、恐懼訴求

以恐懼為訴求的說服技術是最常用的一種方式。勸服者會說：「如果你不照我所建議的方式去做，你將會面臨這些可怕的結果」，例如高中的駕駛訓練課程中，以可怕的車禍場景嚇阻危險的駕駛行為(例如酒後駕駛)，戒菸的廣告訊息中可能運用肺癌的恐怖來展現；口腔衛生的相關訊息以描繪牙齒疾病的毀壞來顯示等等。這些以恐懼為訴求的方式已經被廣泛地研究。然而，其中最核心的問題是，要用多少強度的恐懼訴求才好，愈強愈好嗎？反之亦然嗎？還是其中並無太多的不同。

DEFINING FEAR APPEAL VARIATlONS

In thinking about this research question, it is crucial to be clear about how "strong" and "weak" (or "high" and "low") fear appeals are to be defined. Unfortunately, there are two fundamen­tally different - and easily confused - ways of conceiving of the variation in fear appeals. One way of defining variations in the strength of fear appeals is by reference to the properties of the communication. That is, a "high fear appeal message" is one containing explicit, vivid depictions of negative consequences, and a "low fear appeal message" is a tamer, toned-down version. Notice, however, that this way of defining fear appeal variations makes no reference to the actual arousal of fear in the audience. By this definition, a "high fear appeal message" and a "low fear appeal message" might evoke the same degree of fear.

(一)界定恐懼訴求的變項

思考這個研究問題，清楚的釐定恐懼訴求程度的強弱(或是高低)是非常重要的；不幸的，共有兩種完全不同但極容易搞混構成恐懼訴求的變項。一種定義變項在恐懼訴求強度的方式是藉由參考溝通介面，高程度的恐懼訴求就是將負面的後果清楚的描述出來，低程度的恐懼訴求則是以溫和低調的方式，這一類恐懼訴求變項的定義方式，並不在於對照真正喚起受眾的恐懼；根據這樣的定義，高程度的恐懼訴求與低程度的恐懼訴求都可能引起相同程度的恐懼感受。

The second way of defining fear appeal variations is by reference to the degree of fear aroused in the audience. In this conception, a "high fear appeal message" is a message that evokes comparatively greater fear or anxiety in receivers, and a "low fear appeal message" is one that elicits relatively less fear. (165)This way of defining fear appeal variations makes no reference to the intrinsic characteristics of the message, but instead uses the degree of aroused fear as the index for fear appeal variations.

第二種定義恐懼訴求變項的方式，則是藉由參考受眾所產生的程度而定；以此概念，所謂高程度的恐懼訴求是指引起受眾高度的恐懼或焦慮，而低程度的恐懼訴求則是引起較少的恐懼。這樣的定義方式，並不以參照訊息本身的特性，而是以該訊息所引起的恐懼程度以為恐懼訴求變項的尺度。

Obviously, these two ways of conceiving of "high versus low" fear appeals are quite different. A message might be a "high fear appeal" by the first definition (because it has lots of gruesome content) but not by the second (if it fails actually to arouse fear). And, equally obviously, there's a potential for great confusion in thinking about a research question such as "Are high fear appeals more effective than low fear appeals?" This question might be interpreted either as "Are messages with more gruesome contents more effective than those with less gruesome contents?" or as "Are messages that arouse greater fear more effective than those arousing less fear?" As it turns out, the research evidence clearly indicates the importance of distinguishing these questions.

顯然這兩種設想“高與低”恐懼訴求的方式相當的不同，一份訊息可能在第一種定義屬於強度的恐懼訴求(因為它含有很多恐怖的內容)，在第二種定義則不然(因為它無法真正引起受眾的恐懼)。另外同樣明顯的是，有一種關於一個研究問題的潛在巨大混亂，那就是高度的恐懼訴求比低度的恐懼訴求較具說服力嗎？這個問題可以說成“是否含有較多恐怖內容的訊息比起含有較少恐怖內容的訊息較具說服力？”或“是否引起較多恐懼心理的訊息比引起較少恐懼心理的訊息較具說服力？”研究證據清楚地顯示區分這些問題的重要性。

THE RESEARCH EVIDENCE

The research concerning fear appeals suggests two general conclusions. The first is that message material that is intended to a high level of fear or anxiety in receivers may or may not do so. Boster and Mongeau's (1984) review of fear appeal research found that message manipulations aimed at inducing different levels of fear just weren't very effective at doing so (across 40 studies, the average correlation between fear manipulations and perceived fear was only .36). That is to say, it's not easy to manipulate the level of fear experienced by an audience. Even experienced researchers working with carefully controlled experimental materials have found it difficult to induce the intended degrees of fear dependably. Notice, then: One can compose a message that is very carefully designed to arouse fear and anxiety - a message containing all sorts of material intended to create  fear and yet that message may fail to do so.

(二)研究證據

關於恐懼訴求的研究顯示兩個通論，第一個是想要使受眾在感到高度恐懼或沮喪的訊息內容，並不一定可以獲得想要的結果。Boster and Mongeau's (1984)重新審視恐懼訴求的相關研究，發現想要藉由操縱訊息內容來引發受眾不同程度的恐懼情緒是不容易奏效的(共40份研究，發現恐懼訴求的操縱與所引發受眾的恐懼間平均相關係數為0.36)。也就是說，想要操控受眾的恐懼感受並不是件容易的事，即使是一位有經驗的研究者，在控制良好的實驗環境下，都很難有相當可信度的操控恐懼級數。重點就是我們很容易去精心設計一份充滿引發恐懼、焦慮感受的訊息，但訊息卻無法的任意引發受眾的恐懼感受。

The second conclusion is that message material that does induce greater fear or anxiety will, as a rule, enhance the effectiveness of the message. Two meta-analytic reviews of fear appeal research have concluded that higher levels of induced fear are associated with greater persuasive effectiveness (Boster & Mongeau, 1984; Sutton, 1982; for an earlier review, see Higbee, 1969); that is, receivers who report greater fear/anxiety following the persuasive message are also more persuaded by the message.

第二個結論就是達到引發受眾產生大量恐懼與焦慮感受的訊息，通常可以強化訊息的效果，兩份對於恐懼訴求研究的資料分析，發現引發愈高度的恐懼感與更強力的說服效果相關聯。也就是說受眾回應隨著說服性訊息而產生的高度恐懼與焦慮，也較能被該訊息所說服。

Taken together, these two general conclusions should make it clear that a message offering explicit, vivid depictions of undesirable possible outcomes might or might not be especially effective in persuading. If such a message fails to arouse much fear, then the message may not have much impact. Persuaders should not assume that a message with intense, gruesome contents will auto­matically generate substantial fear in the audience. On the contrary, the research evidence indicates that it's quite difficult to manipulate fear dependably through persuasive messages. But messages that do successfully arouse relatively greater fear are likely to be more persuasive than those that arouse less fear.

綜合來說，這兩個通論使讓我們更加明瞭，一個提供清晰、生動但並不被想要的可能結果的描述，不見得必然可以更具說服力；當這樣的訊息不能引發大量恐懼，這訊息也就不是那麼有影響力。說服者不能假定強烈的、恐怖的訊息內容必能引發受眾真正的恐懼；相反的研究證據指出，僅靠說服性訊息去操控恐懼是相當困難的，但是成功引起受眾較多恐懼感的訊息比起引起受眾較少恐懼感的訊息來的有說服性。

A CURVILINEAR RELATIONSHIP

Earlier summaries of fear appeal research sometimes suggested that the relationship of fear appeal level to persuasive effectiveness is a curvilinear one, with the greatest effectiveness coming at some moderate level of fear appeal (e.g., Janis, 1967). And intuitively, this might seem to have some appeal; after all, one might think, can't a persuader "go too far" in using fear appeals, and thereby experience reduced effectiveness? But it is important to be clear about what might be meant by "going too far" in using fear appeals. If by "going too far" is meant "arousing too much fear," then the answer suggested by the research evidence is that it's not likely that a persuader can arouse too much fear. As discussed above, increases in reported fear are reliably associated with increases in persuasive effectiveness. The evidence to date gives support for the idea that persuasive effectiveness diminishes at high levels of aroused fear (for general reviews of the relevant work, see Boster & Mongeau, 1984; Sutton, 1982; for a useful critical discus­sion of the curvilinear hypothesis, see Leventhal, 1970, pp. 160-168).

(三)曲線關係

稍早對於恐怖訴求的摘要，建議恐怖訴求與說服效果是呈曲線模式，在適當等級的恐懼訴求中可獲得最好的效果。不過另ㄧ個值得思考的問題就是，會不會因為勸服者過度運用恐懼訴求，使得恐懼訴求失去了原有的效能；但是要如何界定「過度運用」恐懼訴求是非常重要的，如果是指的是引起過多的恐懼情緒，則研究證據顯示勸服者未必能引起太多的恐懼，總之增加恐懼可以增加說服效果，但最新的證據顯示在引起高度的恐怖情緒後，說服效果會減少。

But if by "going too far" in. using fear appeals is meant using message material that is too graphic, too explicit, too gory, then perhaps a persuader can overdo it. But notice that this is entirely consistent with the finding that increases in aroused fear are associated with increases in effectiveness. After all, a message that is too gruesome may end up not arousing very much fear (perhaps because the audience doesn't pay much attention). In such a case, the persuader might need to tone down the message material in order to increase the level of fear produced (and thereby increase the message's effectiveness). Remember: The research evidence indicates that it's not easy to manipulate the level of fear experienced by receivers; we should not be surprised if it should turn out that in some cases a great deal of graphic content doesn't arouse all that much fear.

但是如果「過度運用」恐懼訴求指的是運用過多真實、血腥、恐怖的訊息內容，則說服者就可能做的太過頭。但是，這和引起越多的恐懼就有越多的效果的發現完全一致；總之，過於可怕的訊息內容反而可能無法引起受眾的恐懼(或許是受眾無法專注於該訊息內容)。如此，為了增加受眾所產生的恐懼，說服者可能需要降低訊息內容的恐怖(如此方能增加訊息的說服程度)。記住，研究證據指出要操控受眾的恐懼體驗是非常不易的，所以當有些非常恐怖的訊息內容無法相對引起受眾的恐懼，就不足為奇了。

EXPLAINING FEAR APPEAL EFFECTS

Why is greater aroused fear associ­ated with greater persuasive effectiveness? That is, what is the explanation for this observed effect? At present there are a number of competing explanations, and it's not clear which is the best (for discussion of some of the alternatives, see Beck & Frankel, 1981; Leventhal, Safer, & Panagis, 1983; Sutton, 1982).

But it should not be assumed that an explanation of these findings must give a central causal role to fear (that is, to the emotional reaction evoked by the message); instead, an explanation might emphasize cognitive reactions to the message. For instance, a given message might induce more fear and anxiety in receivers - but it may also lead receivers to believe that the fearful conse­quences are more severe (more harmful, more noxious, more disadvantageous) than they had previously believed. And it might be that the real force at work behind the message's effectiveness is the change in those beliefs, not the arousal of fear. Fear might arise as a by-product of the persuasion process ("I now believe the consequences of poor dental hygiene are much more severe than I did before, and this makes me more anxious about this than I was before"), but the cognitive changes - not the emotional ones - might actually explain the message's effectiveness. From this point of view, then, greater aroused fear will (167) be associated with greater persuasive effectiveness--not because greater fear causes greater effectiveness, but because both fear and effectiveness are caused by the same underlying factor (namely, the cognitive reactions to the message).

(四)解釋恐怖訴求的效果

為什麼高度的恐怖訴求會產生高度的說服效果？這就是所觀察到的效果之解釋。目前，有一些相對的解釋，卻仍不能確定何者為最佳解釋。但是不應假設那些發現的解釋，必須是引發恐懼的中介角色(因為對於訊息內容的情緒性反應)；反而是一種可以強調對於訊息的認知反映的解釋。例如一個賦予的訊息可能引發受眾的恐懼與焦慮，但它也有可能導引受眾比以前相信可怕的後果比起原先所預期的更嚴重、有害與不利。也許改變他們的信念才是訊息效果背後真實的力量，而不是引發恐懼。所引起的恐懼只是說服過程的副產品(比起從前我現在比較相信，不良的口腔衛生是非常嚴重的，而這也使得我對於這件事更加感到焦慮)，但是認知的改變，並非情緒的轉變，可以確切的解釋訊息的效果。從這個觀點看來，引起愈多的恐懼與愈多的說服效果相關，不是因為大量的恐懼產生效果，而是因為恐懼和效果都是被稱為“對於訊息的認知反應”所導致而來。

In fact, one central way in which the various fear appeal explanations differ is precisely whether the key factor at work is taken to be an emotional reaction (fear) or a cognitive reaction (e.g., a judgment of the seriousness of the problem being depicted). Some theoretical accounts emphasize the causal role of fear (e.g., Hovland, Janis, & Kelley, 1953), whereas others place much greater emphasis on cognitive processes (e.g., Beck & Frankel, 1981; Rogers, 1975;Sutton,1982). In sorting out these various explanations, one important general question is whether fear has some effect on persuasive outcomes over and above the effects of accompanying cognitive changes. If fear does have such effects, then presumably a good explanation will need to give some causal role to the emotional reaction; if fear does not have such effects, then a purely cognitive explanation might be sufficient. On this issue, research evidence is only begin­ning to be acquired (see, e.g., Sutton & Eiser, 1984; Sutton & Hallett, 1988, 1989), and hence conclusions are probably premature. In any case, one ought not assume that the observed covariation between induced fear and effectiveness means that induced fear causes effectiveness.

事實上，在中央途徑中不同恐懼訴求的解釋之中，有一個重要的因素就是迫使受眾產生情緒(恐懼)或認知上的反應(例如，判斷問題的嚴重性)，一些理論強調恐懼的觸發角色，其它地方更強調認知過程。在挑選出來這些不同的解釋中，雖然在其他地方更強調認知的過程，一個非常重要的問題，對於超出伴隨認知改變而來的說服結果，恐懼是否真的有效果；如果恐懼真的有這樣的效果，就可以假設需要對於情緒上的反應給予觸發的角色一個合理的解釋；如果恐懼沒有這樣的效果，那麼一個單純認知上的解釋也許就足夠了。在這個議題上，研究證據才剛開始獲得，因此這樣的結論並不夠成熟。在任何的情況下，不應該假設引發恐懼與效果間的共變數就代表引發恐懼就形成效果。

EXMPLES VERSUS STATISTICAL SUMMARIES

The research question here concerns the relative persuasive impact of two different forms of information: the example (or case history), which describes some event or object in detail, and the statistical summary, which provides a numerical summary of a large number of events or objects (in the form of averages, percentages, and so on). What is often compared, then, is information about the experience of one (or a very few) individuals and equivalent, summa­rize statistical information about the experiences of many individuals.

(五)舉例與統計摘要

這裡的研究問題探討兩種不同形式資訊的說服衝擊，關於詳細描述一些事物以及統計上摘要的例子，提供了一些大量事件或物體的數據摘要(以平均數、百分率等方式呈現)，最常拿來比較的就是個人(或少數幾個人)及對方的經驗，概述許多個人經驗的統計數據。

For example, Koballa (1986) provided preservice high school teachers with favorable information about a particular type of science curriculum in one condition the information was presented as the report of a single teacher who had used the curriculum; the teacher discussed how much more interested the students were, how much more the students learned, and how student performance improved in related areas (such as math and writing). In the other condition, the very same points were made (about students being more interested, learning more, improving in related areas), but the information was presented as a statistical summary of the findings of a dozen or so different uses of this curriculum.

例如，Koballa (1986)提供孰悉自然科學課程特殊形式的高中老師，這類資訊以使用這課程老師的報告來呈現，老師描述那些學生對於課程的興趣愈多，那些學生的收穫就愈多，以及那些學生就可以在該領域進步多少(例如數學及作文)。在其它的情形下，產生同樣的結論(關於學生愈有興趣就學的愈多，也就在相關的領域得到更多的進步)，但是這些資訊可以用在這些課程上所發現既多且不同用途的統計摘要來呈現。

KobalIa (1986) found that the case study report was much more persuasive than the statistical summary report (even though, obviously, the statistical summary was based on the experience of many different teachers, not just one). And indeed this is the general finding in this line of research: Examples or case histories are more influential than statistical information or other data summa­ries (for a review, see S. E. Taylor & Thompson, 1982).

Koballa(1986)發現這份個案研究報告比起統計式的摘要更加具有說服力(即使如此，顯然的這些統計數據是根據許多不同老師的經驗，而非單單來自一個老師而已)，而且在這一類的研究中都有同樣的情形，案例比起統計數據或資料摘要更具說服力。