Persuasive evidence in India: an investigation of the impact of evidence types and evidence quality.

AuthorHornikx, Jos
PositionReport

For more than fifty years, researchers have been interested in the kinds of data that are most effective in supporting a claim (Reinard, 1988). When data are used as proof for claims, they can be considered evidence (cf. Reynolds & Reynolds, 2002). Evidence has mostly been studied for claims that describe (future) events: factual claims or probability claims. Generally speaking, researchers have investigated the persuasiveness of evidence types by comparing claims (in a longer text or in isolation) supported by one or another type of evidence: anecdotal, statistical, causal, or expert evidence. Anecdotal, or story, evidence (Baesler, 1997) usually consists of an individual who has experienced the event described in the claim. Statistical evidence provides numerical data based on a large number of individuals who have experienced the event. Causal evidence provides a causal explanation of the event described, and, finally, expert evidence consists of a reliable and trustworthy expert who underscores the claim.

Anecdotal and statistical evidence have been most frequently studied. In Slater and Rouner (1996), for instance, anecdotal evidence consisted of a person who had experienced the problems described in the claim because of his alcohol consumption (namely, economic and career harm). An example of statistical evidence was that 25% of men who occasionally drank suffer from the problems described. Slater and Rouner (1996) found statistical evidence to be more effective at supporting claims than anecdotal evidence. In other studies, anecdotal and statistical evidence were found to be equally persuasive (e.g., Baesler, 1997), or anecdotal evidence was found to be most persuasive (e.g., Koballa, 1986). Research findings on the persuasiveness of these two types of evidence have been equivocal, partly because of large variations in manipulations of evidence (Kellermann, 1980). In order to summarize extant research on a given domain, a meta-analysis is well suited (e.g., Allen, 2009; Lipsey & Wilson, 2001). Allen and Preiss (1997), using a meta-analysis, summarized the empirical studies that addressed the question whether statistical evidence is more or less persuasive than anecdotal evidence in support of claims. Based on 19 comparisons of statistical and anecdotal evidence, Allen and Preiss concluded that statistical evidence was generally more persuasive than anecdotal evidence. In a narrative review, which included the findings of more recent experiments, Hornikx (2005) supports this conclusion.

In recent studies with multiple message designs, in which participants judged a number of claims that were supported by different evidence types, findings are fairly homogeneous, and corroborate Allen and Preiss' (1997) conclusion: statistical evidence is more persuasive than anecdotal evidence (see Hoeken & Hustinx, 2003, 2009; Hornikx & Hoeken, 2007, Study 1). In Hoeken and Hustinx (2009, Study 1), for instance, participants judged the probability of 20 claims, which were supported by either statistical or anecdotal evidence. In general, participants judged claims that were supported by statistical evidence as more probable than claims supported by anecdotal evidence. Two other studies used the same research design to compare these two types of evidence with causal and expert evidence (Hoeken & Hustinx, 2003; Hornikx & Hoeken, 2007). Findings show that causal and expert evidence were more persuasive than anecdotal evidence (Hoeken & Hustinx, 2003; Hornikx & Hoeken, 2007, Study 1). In Hornikx and Hoeken (2007, Study 1), statistical evidence was not only more persuasive than anecdotal evidence, but also more persuasive than causal and expert evidence.

PERSUASIVE EVIDENCE IN WESTERN CULTURES

Allen and Preiss (1997) observed that the experiments they included in their meta-analysis were all conducted with Western participants. They therefore suggest that future research should consider other cultures: "Whether other cultures with different expectations for forms of proof would reflect the same outcomes is unknown" (p. 129). In the early stages of research on evidence, researchers already have asked whether the persuasiveness of evidence may differ from culture to culture (McCroskey, 1969). Although researchers have regularly emphasized the relevance of this question (Greene & Brinn, 2003; Reynolds & Reynolds, 2002), it has not received a developed answer yet.

Only a few studies have been conducted on culture and evidence, all of which have compared Western cultures. First, Parrott, Silk, Dorgan, Condit, and Harris (2005) examined the perceived persuasiveness of statistical evidence in verbal or visual format among African Americans and European Americans. The two cultural groups perceived statistical evidence with a visual representation as equally persuasive as statistical evidence with a verbal representation. Second, Hornikx (2008) gave Dutch and French participants eight different claims, each followed by a list of four types of evidence. The participants ranked these different evidence types in terms of how persuasive they expected the evidence to be for others, starting with the most persuasive on rank 1, and so on. For the two cultural groups, the pattern was similar: participants expected statistical evidence to be most persuasive for others, followed by expert, causal, and anecdotal evidence. Hornikx and Hoeken (2007, Study 1), finally, investigated the actual persuasiveness of anecdotal, statistical, causal, and expert evidence in 20 claims that they presented to participants in France and the Netherlands. For both cultural groups, statistical, causal, and expert evidence were found to be more persuasive than anecdotal evidence. For the Dutch participants only, statistical evidence was also more persuasive than expert and causal evidence. Hornikx and Hoeken (2007, Study 2) also manipulated the quality of evidence. Evidence quality depends on the underlying argumentation scheme that is present when evidence supports claims. Evidence has a high quality (and is normatively strong) when it respects the norms associated with the argumentation scheme. For instance, for the argument by authority, a person's credibility, reliability, and relevance of the field of expertise in relation to the topic of the claim are important (Walton, 1997). Hornikx and Hoeken (2007, Study 2) manipulated expert evidence to have a high quality when the expert's position ensured credibility and reliability, and when her or his field of expertise was relevant to the claim's topic. Expert evidence had a low quality when the expert's field of expertise was irrelevant to the topic of the claim. The quality of statistical evidence was manipulated by reducing the number of cases in the evidence, which is important for the underlying argument by generalization (cf. Garssen, 1997). Hornikx and Hoeken (2007, Study 2) showed that Dutch participants, but not the French participants, were sensitive to the quality of both expert and statistical evidence. For the Dutch participants, normatively strong statistical and normatively strong expert evidence were more persuasive than normatively weak statistical and normatively weak expert evidence.

EVIDENCE AND WESTERN-EASTERN SYSTEMS OF THOUGHT

The studies comparing different Western cultures indicate that culture can play a role in the persuasiveness of different evidence types. The insights that these studies provide are limited in scope because, as Allen and Preiss (1997) observed, the empirical research on evidence has been conducted in the United States and Europe only. It is unknown how persuasive the different types of evidence are in cultures other than Western cultures, such as cultures in the Eastern part of the world. The comparison between so-called Western and Eastern cultures has been very popular in cross-cultural research in general (e.g., Hofstede, 2001; Nisbett, 2003; Oyserman, Coon, & Kemmelmeier, 2002), and in argumentation research in particular (e.g., Combs, 2004; Gillon, 2008; Oetke, 1996). It should be noted that researchers emphasize that, within these two broad categories, there exist important differences. For instance, Varnum, Grossmann, Katunar, Nisbett, and Kitayama (2008) demonstrated that the cognitive style of Americans and West Europeans was different from Central and East Europeans, even though all four are categorized as Western. Knight and Nisbett (2007) even showed that significant cognitive differences can exist within one culture, depending on a person's social class. Nevertheless, specific differences can consistently be observed between Western cultures on the one hand and Eastern cultures on the other.

One of the frameworks comparing these two broad cultural groups is the psychological notion of Western and Eastern systems of thought (Nisbett, 2003; Nisbett, Peng, Choi, & Norenzayan, 2001). Western cultures, such as those in North America and Europe, are heavily influenced by ancient Grecian cultures (Nisbett, 2003)...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT