Dimensions of Perceived Information Quality and Its Role in Information Processing

ABSTRACT - The extent to which consumers have difficulty in making informed judgments on which to make product evaluations is tested on a number of key dimensions of food labelling. Using various dimensions of perceived information quality, it is shown that consumers find much information confusing and rely on peripheral processing. All businesses should be concerned about the clarity of information and presentation of their ingredients. The tendency for association between different dimensions of information quality suggests some manufacturers could improve their brand positionings.



Citation:

Mark A P Davies (1995) ,"Dimensions of Perceived Information Quality and Its Role in Information Processing", in E - European Advances in Consumer Research Volume 2, eds. Flemming Hansen, Provo, UT : Association for Consumer Research, Pages: 204-208.

European Advances in Consumer Research Volume 2, 1995      Pages 204-208

DIMENSIONS OF PERCEIVED INFORMATION QUALITY AND ITS ROLE IN INFORMATION PROCESSING

Mark A P Davies, Lougborough University

ABSTRACT -

The extent to which consumers have difficulty in making informed judgments on which to make product evaluations is tested on a number of key dimensions of food labelling. Using various dimensions of perceived information quality, it is shown that consumers find much information confusing and rely on peripheral processing. All businesses should be concerned about the clarity of information and presentation of their ingredients. The tendency for association between different dimensions of information quality suggests some manufacturers could improve their brand positionings.

INTRODUCTION

The role of information in decision choice

Marketing information is recognised to be processed by how consumers interpret information and transform it in to knowledge or meaning, and how they integrate this with the stored knowledge in memory (as elaboration) to form judgments about products and brands, or develop appropriate behaviour (Chestnut and Jacoby, 1977; McGuire, 1976; Preston, 1982, Olson, 1993).

Relevance of information search

According to Bettman and Kakkar (1977), the format in which information is presented determines the extent to which information is acquired and processed, ultimately affecting how brands are evaluated. More specifically, Petty and Cacioppo, (1984); Areni and Lutz, (1988), explain that the details of argument quality in promotional claims will only be assessed if there is both motivation and ability to do so. In other instances, peripheral cues or short-cut rules will be used to make a judgment.

IMPORTANCE OF FOOD LABELLING INFORMATION

Marketers have considerable control over many aspects of the consumer's information environment. Two thirds of purchases are decided in-store (Olson,1993), with packaging often serving as the last opportunity for using product information for guiding purchase behaviour. Therefore, an important source of information for many consumer products is the label, with food representing a significant part of household purchases. Labelling covers a vast amount of information, including ingredients and additives, nutritional content, marketing claims, quality assurance schemes, country of origin marks, instructions in storage and cooking and product descriptions. A survey by the National Food Processor's Association revealed that 79% of consumers read labels on food products before buying for the first time, and over 1/3 of consumers always read ingredients or nutritional information (Mueller, 1991).

One particular aspect of labelling is the issue of how additives should be displayed and presented, together with associated food claims about them, summarised by Davies and Wright (1994). They record a number of published findings which suggest that consumers have difficulty in making informed decisions from the data presented to them on labels. For example, E numbers may be natural or unnatural, but to many consumers they are only perceived as unnatural (The Grocer, 1988). Many food claims such as 'high fibre' or 'low fat' are misleading, according to the Coronary Prevention Group (Hoggan, 1992). Davies and Wright (1994) purport that an inability to understand information on food labelling may contribute significantly to the discrepancy between concern over additives (as measured by attitudinal surveys) and contradictory purchase behaviour. Evidence from a Taylor Nelson food panel (1990), suggests the demand for more informative food labelling is not simply confined to niche markets, but to a significant number of consumers who lack the confidence in making an informed decision.

This paper attempts to compliment the published literature, in applying the ELM to labelling information. Research is used to examine the extent to which consumers have neither the ability nor motivation to interpret labelling information to make an informed judgment. Several dimensions of ability are measured. These are consumer beliefs about the ease in interpreting labelling information measured by the clarity, readability, ease in finding information, and the ease to which the consumer could position the brand on purity and healthiness. Beliefs about the relevance of information, as a dimension of motivation, is also measured. Collectively, these beliefs are subsequently referred to as a measure of perceived information quality, (PIQ). In the absence of making informed judgments, the types of peripheral cues used are noted. Recommendations and conclusions are presented for brand management.

METHODOLOGY

The chosen sample were split in to smaller groups and were invited to participate in a series of in-hall tests, in which they had an opportunity for viewing various brands on display. Care was taken to ensure there was no collusion with other sample members at this stage.

A questionnaire was designed to test their level of agreement with various dimensions of PIQ, using a likert scale anchored by 1=agree strongly, to 5=disagree strongly. Pilot testing had been conducted to refine the wording of the questions. The brands tested were labelled A-G. Specifically, sample members rated each brand, based on their beliefs, towards agreeing with a series of attitudinal statements to measure the following dimensions of PIQ:

(A) the (e)ase with which each brand can be (d)etermined (p)ure or not, from initial (p)erceptions, (EDPP)

(B) clarity of labelling information (how easy to understand)

(C) readibility of labelling (how easy to read)

(D) relevance of labelling (how relevant is it for my needs)

(E) ease of finding labelling information

(F) the ease with which each brand can be (d)etermined (p)ure or not from the list of (i)ngredients, (EDPI)

(G) the (e)ase with which each brand can be (d)etermined how (h)ealthy it is, from the (c)laims made on the labelling, including nutritional information, (EDHC).

These dimensions, derived from several earlier focus groups of a similar demographic mix, appeared to affect how prominently labelling was used to make informed judgments about brands. They were considered to be important dimensions of PIQ.

Beliefs about the ease with which the data could determine brand purity, were measured both holistically, i.e., from initial perceptions, without resorting to examining each brand in detail (referred to as EDPP), and from the list of ingredients; (EDPI). Holistic perceptions of brand purity are based on the appearance of the brand, including prominent brand claims but not the detailed ingredients. Beliefs were also measured for the ease of determination of healthiness from the claims made (EDHC).

The sampling design was a multi-stage cluster area technique. Firstly, an urban council area was selected, considered to be demographically representative of the national average. Electoral wards were then selected within the council area, and then household units within these selected electoral wards. Ten wards were selected from the council. A quota for each ward was set, determined by the relative population size of each ward. Originally 399 people were contacted in order to obtain a final sample size of 200 who agreed to participate in the consumer research. Actual households were chosen randomly, by allocating a number to a list of addresses by ward, and generating actual members from random number tables. The design offered the advantages of convenience without the intentions of selection bias.

The person responsible for the main household shopping duties was chosen as the sample member per household. This is because they are most likely to be experienced purchasers of the product category, and part of the target audience. This would tend to underscore any problems the sample met in comprehending and interpreting labelling information, since it would be assumed they should hold more highly developed knowledge structures.

In choosing a product category, small price differentials between brands were ideal, to reduce the effects of the price cue on attitudes towards information provision. A fast moving consumer good was therefore chosen. The product category squashes met this criterion, and seven brands were chosen from several multiple chain stores. Each brand name on the surface label was removed and appropriately retouched with the background colour of the label to reduce potential bias of brand effects between major and minor brands. For example, conceptual input data (such as brand names) might encourage additional inferences to be made (e.g., perhaps based on prior experiences with similar brands) beyond the perceptual information transmitted (Cherian and Jones, 1991). Removing the brand name reduces the chances that consumers will process information conceptually rather than perceptually.

Data analysis was conducted by the Minitab statistical package. Frequency tables of their beliefs towards agreeing with each attitudinal statement per brand were firstly constructed for each rating 1-5, based on the likert scale. The responses for ratings 1 and 2; and 4 and 5 were then pooled to obtain percentages of those who agree and those who disagree with each statement. The degree of association between pairs of attitudinal statements was then examined, through cross tabulations and the chi-squared statistic. This was to identify any strong links in levels of agreement between attitudinal statements. Strong links indicate the potential of such data in reinforcing brand positioning.

A further study aimed to interpret the reasoning behind these results. Two focus groups were conducted, drawn from the original sample, consisting of 8 and 7 respondents. Throughout the discussions, the same brands of squash were on display. The discussions were transcribed for post-session analysis. It was felt there was no need for further focus groups, on the grounds that findings in the second focus group began to replicate those in the first.

RESULTS AND ANALYSIS

Average levels of agreement per attitudinal statement per brand were computed. The highest and lowest average brand scores per attitudinal statement are graphically presented in Figure 1, together with the average level of agreement per statement for all brands (AV). Percentages of agreement are shown on a scale of 1.0 to zero. Analysis of branding information measured both the spread of average agreement between each brand (shown by the length of each line in Figure 1), and the average for all brands (AV) to assess how typical or atypical an overall brand is, based on its average scores. Thus a long line suggests a greater degree of variability between brands, in which a favourable rating offers more scope for gaining competitive advantage, ceteris paribus. A short line suggests that ratings are clumped close to the average of all brands, indicating less scope for competitive advantage. What this figure does not show is the variation in individual responses, since average responses towards each attitudinal statement are more important.

The results of consumer beliefs show that:

In examining statement A, purity cannot be easily determined from the appearance of the product alone for most brands, since the average of all brands (AV) is 0.52, with the lowest brand scoring 0.39. Focus groups identified cues of purity by the colour of the orange, the design of the label and whether selective claims were present or not (e.g., 'contains no artificial colours') rather than by an informed judgment of the contents. In many cases, due to widespread practice of using selective claims, the lack of them were rationalised as a signal of unfavourability. Respondents appear to be using the 'proximity' principle, in the absence of more tangible cues of purity. For instance, some respondents associated a rich orange colour as purer than a lemon or pineapple-looking colour for some orange squashes. Another indicator of purity was the materials used in presentation. Glass was considered superior in conveying a fresh colour, enhancing the image of purity, compared to plastic packaging. Another factor was the graphical detail of the surface label. One brand was presented with a water drop over a picture of an orange, which apparently gave the impression of fresh oranges. Consistent with figure and ground perceptual theory (Engel, Blackwell and Miniard, 1993), most respondents thought a dark background, with an orange foreground to be most appealing.

In examining statement (B), the clarity of labelling information for all brands was poor, with the average brand scoring only 0.21, and the highest only 0.30. Information on labelling was difficult to understand because it appeared to be presented too technically, or was incomplete. Although the main headings of the ingredients were clear to understand, there was general confusion with chemical names (e.g., those representing sugar or sugar substitutes). Also, the terms 'kcal' or 'kj' were not understood in relation to the number of calories. Several claims were considered not specific enough, such as 'low in calories'. One respondent suggested that claims such as 55% orange juice' led them to ponder on the contents of the remaining 45%. More women than men stressed the need for a simplified , standardised labelling format for all products, which would make comparisons easier. For example, date marks were not standardised, with some brands presented with 'use by' dates, others by 'best before' dates, causing difficulties in making strict comparisons.

Although most brands scored reasonably on readability (as indicated by the mean brand value of 0.74, statement C), one brand scored poorly, at 0.325. According to the focus group discussions, contributions to reading difficulty included size of print (too small), and cluttered information. The visual impact of some designs was believed to be better for some brands, accounting for the large spread in overall brand score averages. One brand presented its nutritional information vertically, when in situ on the shelf, making it significantly harder to read.

For statement D, much information appeared to be irrelevant for their needs with the average brand scoring only 0.37. In the focus groups, respondents were asked what aspects of information were most relevant to them. Product claims (e.g., 'sugar free', 'low calorie', 'no added sugar'), were more likely to influence initial purchasing considerations but not necessarily final preferences. Claims were often incomplete or selective and believed to be less than informative. The chi-squared statistic indicated a significant association between the levels of aggreement for clarity and that for relevance (except for brand E, with c2calc=1.30, c2crit, 0.05=3.84), suggesting that irrelevance may be due, in part, because consumers cannot make sense of such information. This tends to support the peripheral route to persuasion, under the ELM theory, in which inability to process information affects motivation (Petty and Cacioppo, 1984).

FIGURE 1

HIGHEST, LOWEST AND AVERAGE (A) BRAND SCORES

For statement E, there is a great variation in ratings between the highest and lowest scoring brand (0.87 and 0.23 respectively). Participants suggested that typeface should be easily seen from a normal shelf position, without resorting to having to pick it up and examine it. Although most brands adhered to this simple rule, one general criticism applied to storage instructions: respondents felt that they should be given a more prominent location on the product.

For statement F, most consumers appear to have some difficulty in discerning purity from the list of ingredients (with the average brand score at 0.38, and the top brand only registering 0.53). Focus group discussions suggested that other sugars (apart from sucrose) were not always declared, which meant that the 'sugar free' claim may be bogus or too vague to be useful. For example, despite one brand claiming to be 'sugar free', it contained saccharin, an artificial sugar, believed to be misleading. This explained why several respondents thought that nutritional information was not a foolproof indicator of healthiness. This is reflected in the results for statement G, in which the average brand score for determining healthiness from the nutritional information is only 0.53.

The chi-squared statistic revealed that for most brands, there was a significant association between the level of agreement with pairs of attitudinal statements. Although the actual chi-squared statistic for each brand is not shown to retain parsimony, the strength of association, determined by the number of brands showing statistical significance, summarised in Figure 2, is represented by the number and thickness of the lines used, according to the key. For each brand, there was a strong association between clarity and EDHC, and a very strong association between EDPI and EDHC. Associations between clarity and EDPI, relevance and EDPI and relevance and EDHC tended to be brand specific.

FIGURE 2

RELATIONSHIP OF ASSOCIATION BETWEEN LEVEL OF AGREEMENT (USING PAIRS OF ATTITUDINAL STATEMENTS)

Focus group discussions suggested that, in the absence of more tangible information, consumers use peripheral cues in the processing of labelling information. Interestingly, brand scores for EDPP outperform those of EDPI, suggesting the strong role of peripheral cues used in brand comprehension and evaluation. Although the association between EDPP and EDHC appears to be generally significant, it is much less strong than for that between EDPI and EDHC. Considering many consumers have difficulty in understanding the ingredients (based on earlier findings) suggests the health positioning of some wholesome brands might be suffering unduly. This is further reinforced by the general significant association between clarity of information and ease of determination of healthiness. Consumers appear to rationalise that if the labelling is unclear, businesses have something to hide, implying an unhealthy brand. Indeed, since the association between clarity and ease of determination of purity by ingredients is brand specific, this would suggest some ingredients lists could be refined, possibly leading to more favorable brand positioning.

DISCUSSION AND RECOMMENDATIONS

This study in the squashes market has demonstrated that the PIQ of labelling presented to consumers could be much improved in order to help guide them to make more informed choices. Although the study would need broadening before generalising on the severity of the problem, several published sources imply its severity. One issue of future importance is to examine the relationship between effort required in determining purity compared to the perceived purity; and effort required in determining healthiness compared to perceived healthiness per brand. In other words, is there a strong relationship between these dimensions of PIQ and favourable positioning? Such a relationship might imply that some manufacturers, who provide impure and unhealthy products, may deliberately conceal or mislead about their wholesomeness. More positively, manufacturers of wholesome products who improve their presentation might encourage consumers to evaluate on the basis of more detailed product information, and so develop a competitive advantage.

Future research could examine how brand effects may influence beliefs towards various dimensions of PIQ. For example, does a well known brand name influence the relevance of information, or facilitate in the determination of purity or healthiness? This might be achieved by setting up a control group, subjected to a similar in-hall test, but with the brand names revealed.

Another aspect is to examine differences in reactions between different demographic profiles. The fact that some individuals varied in their beliefs about the quality of information provided by each brand, indicates that consumers hold different levels of ability, motivation and opportunity, and this may affect directly the way information is processed (Cherian and Jones, 1991). Despite individuality, many distinctive patterns of agreement can be identified from the average consumer responses.

In summary, where the spread of average scores between brands is widest, this offers the greatest scope for poorly rated brands to improve how they use information in their labelling strategies. Whilst only some businesses may need to be concerned about responses to all of the dimensions of PIQ (Figure 1), every business should be concerned about responses to statements B and F (since the best average brand scores for any brand are 30% and 53% respectively). This means that clarity of information and the nature of how ingredients are presented need to be considerably improved across the whole market. Since the sample held some experience with the product category, this is particularly important. The amount of association in levels of agreement between various attitudinal statements (Figure 2) also infers the tendency for high elaboration on many dimensions of PIQ. For instance, ratings attributable to clarity and ease of determination of healthiness may infer several things, and could perhaps be used as proxies for measuring other dimensions. Since several of these attitudinal statements appear to be associated with a number of other dimensions, (Figure 2), the implications for minor improvements in PIQ may bring significant gains in how brands are evaluated. Considering clarity is significantly associated with EDPP and relevance of information, a lack of clarity might feasibly undermine EDPP and relevance, contributing towards a reduction in the overall value of information selection and evaluation in the product positioning process. Conversely, improvements in clarity would be expected to make labelling information more relevant, and probably increase elaboration.

Similarly, EDPI is strongly associated with ease of determining healthiness from the nutritional information, and less strongly with EDPP. If EDPI is a proxy for measuring EDHC, the needs of the increasing health conscious segments espoused by Taylor Nelson (1990) are likely to be unsatisfied, since most brands were rated poorly on EDPI.

CONCLUSION

In testing labelling information on a number of key dimensions of perceived information quality, this study has revealed that there is a wide scope for improvement. The close association between many of these dimensions would also suggest that consumers may have difficulty in elaborating information, and rely instead on peripheral processing on which to make judgments. The results from the focus groups support this, showing that peripheral cues, such as packaging graphics and colours are frequently used as a means of determining the purity of a brand.

Although the prescence of peripheral cues can be advantageous, in saving consumer time in decision-making, manufacturers should be cautioned in using these to disguise their product weaknesses. Since many consumers are concerned about what they eat, designing products to mislead is contrary to good marketing practice in satisfying the consumer. It is suggested that labelling information is pre-tested on consumers, along the dimensions suggested.

With the gradual shift towards EC harmonisation of labelling standards across foreign borders, manufacturers may be encouraged to adopt a more pro-active policy, in providing improvements in the quality of information that help reduce peripheral processing and help improve consumer choice. In addition, more use of communications leading up to point of sale, in terms of educating the public about healthy eating with respect to nutritional, dietary and purity claims would increase involvement at point of sale, and improve the contributory value of labelling in brand choice. This is supported by the apparent lack of relevance of much current labelling information, despite consumer concern over what is eaten, and a desire to be better informed.

REFERENCES

Areni, C.S. and R.J. Lutz. (1988), "The role of argument quality in the elaboration likelihood model," Advances in Consumer Research , 15, 197-201.

Bettman, J.R. and Kakkar, P. (1977), "Effects of Information Presentation Format of Consumer Information Acquisition Strategies, Journal of Consumer Research, 3, 233-240.

Cherian, J. and M. Jones. (1991), "Some processes in brand categorizing: Why one person's noise is another person's music," Advances in Consumer Research, 18, 77-83.

Chestnut, R. W. and J. Jacoby. (1977), "Consumer information processing: Emerging theory and findings," in Consumer and Industrial Buying Behavior, eds., A. Woodside, J.N. Sheth and P.D. Bennett, New York: Elsevier-North Holland, 119-133.

Davies, M.A.P. and L. T. Wright. (1994), "The importance of labelling examined in food marketing", European Journal of Marketing, 28 (2) , 57-66.

Engel, J.F., R.D. Blackwell and P.W. Miniard. (1993), Consumer Behavior, Orlando, Florida: Dryden.

Hoggan, K. (1992), "Health charity attacks misleading food claims," Marketing (May), "The week" page.

McGuire, W.J., (1976), "The internal psychological factors influencing consumer choice," Journal of Consumer Research, (March), 302-319.

Mueller, W. (1991), "Who reads the label?," American Demographics,(Jan), p. 36.

Olson, P. (1993), Consumer behaviour and marketing strategy, Boston: Irwin.

Petty, R.E. and J.T. Cacioppo. (1984), "Source factors and the elaboration likelihood model of persuasion," Advances in Consumer Research, 11, 668-672.

Preston, I.L. (1982), "The association model of the advertising communication process," Journal of Advertising, 2, 3-15.

Taylor Nelson. (1990), "A family food panel special report: Healthy eating and environmental concern," Epsom, Surrey: Taylor Nelson, 16-27.

The Grocer. (1988), "Consumers want to buy what comes naturally," (11 Oct), 59-60.

----------------------------------------

Authors

Mark A P Davies, Lougborough University



Volume

E - European Advances in Consumer Research Volume 2 | 1995



Share Proceeding

Featured papers

See More

Featured

Morality Matters in the Marketplace: The Influence of Morally Based Attitudes on Consumer Purchase Intentions

Andrew Luttrell, Ball State University
Jacob Teeny, Ohio State University, USA
Richard Petty, Ohio State University, USA

Read More

Featured

I'm Scared, Want to Listen? Fear's Influence on Self-Disclosure

Anupama Mukund Bharadwaj, University of Washington, USA
Lea Dunn, University of Washington, USA
Joey Hoegg, University of British Columbia, Canada

Read More

Featured

D1. When Intention to Share on Social Media Increases Variety-Seeking: The Role of Self-Enhancement

Jingjing Ma, Peking University
David Dubois, INSEAD, France
Fei Jin, Peking University

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.