THE MULTIPLE DIMENSIONS OF INFORMATION QUALITY
Holmes Miller Muhlenberg College Allentown, PA 18104
Sometimes it pays to go back to basics. This is certainly true when talking about information. Most people agree that information is critical to how we live and work and that information, to be effective, must have "quality". But what is "quality information"?
At first, the concept seems obvious. Quality information meets certain recognized criteria such as "accuracy", "timeliness", "relevance", and "understandability". We know it when we see it. Yet, despite an intuitive grasp of the term, all of us have presented what we thought was quality information to others, only to find that their opinion differed significantly from our own. If the concept is so obvious, then why do so many people disagree on specifics?
The meaning of information quality lies in how the information is perceived and used by its customer. Though absolute attributes are important, it is how those attributes are perceived, now and in the future, that defines information quality. Identifying quality information involves two stages: first, highlighting which attributes are important and second, determining how these attributes affect the customers in question. The following discusses ten attributes of information quality and how they can be used as benchmarks to improve the effectiveness of information systems and to develop information quality strategies for all organizations, particularly for those in the information business.
DIMENSIONS OF INFORMATION QUALITY
The multiple dimensions of information quality are:
The key component for information quality is whether the information addresses its customer's needs. If not, that customer will find the information inadequate regardless of how well the information rates along other dimensions discussed below. This does not mean that irrelevant information to an information customer is of "poor quality". It just indicates that the information is a member of a different information class much in the same way that a luxury car and a sports car are both members of different classes of automobiles. In some cases "poor quality" information may actually be quite good; what is needed is to educate the information's customer so they can understand it and use it.
Accurate information reflects the underlying reality. That quality information should be accurate seems obvious. In practice, information used for different purposes requires various levels of accuracy and it is even possible for information to be too accurate in the sense of being too precise.
Information inaccuracy and related problems occurs in many information systems. The problem is well known and is addressed by information systems professionals through the cycle from systems design to implementation to maintenance. Less well understood is that information can be too accurate when its degree of precision exceeds its customer's processing capability. This can increase information systems cost, become a drain on system credibility and even, through the confusion caused, result in misuse or abandonment.
Timely information is still current. Implicit in this definition is a dynamic process where new information arises to replace the old. Information has a cycle time which depends on how quickly new information can be processed and communicated to its customer.
Information timeliness goes hand in hand with information accuracy. The concept of what is timely is itself constantly changing and being redefined, due to changes in customer perceptions caused by technology and the competitive environment. Today, time based competition and the concomitant reduction in operations cycle times has fueled a demand for evergreen information.
Incomplete information can lead its customer astray. An example of this is the old story of blind travelers encountering an elephant in the road with each, after examining a small section, coming to a different conclusion regarding the object's identity. However, complete information for one person may be incomplete for another. For example, the marketing vice president and the director of research and development for a pharmaceutical company may be both be interested in the clinical trials tests for a new drug, but each may require different levels of detail.
Just as information whose precision exceeds a customer's processing capability may be too accurate, information may also be too complete. The classic example is the attorney dumping boxes of material on an adversary's desk, knowing that it will be impossible to unearth the few relevant facts buried within. The danger in business, are information systems that generate so much information that customers cannot process it all in a timely fashion.
Coherence is how well the information "hangs together" and it consistent with itself. Information can become incoherent through irrelevant details, confusing measures or ambiguous format which can confuse information customers and cause them to not receive or even reject the information's message. Although information can be genuinely incoherent, incoherent information usually indicates an error in accuracy or timeliness.
(Figure 1 Here)
Information format refers to how the information is presented to the customer. Two components of information format are its underlying form and its context for interpretation, which is sometimes referred to as its frame.
The appropriate format for information depends on the information's customer and the information's use. For example, an accountant may prefer twenty pages of numbers to a graphical summary, but may insist of using only multi-color pie charts for his presentation to the senior vice president of sales.
The context with which we view information also is important. For example, when a company benchmarks its performance against industry or functional world class leader, it seeks to give that information context. For years financial analysts have viewed information in context, witness the focus on stock price performance over time, a portfolio manager's performance relative to the S&P 500, or corporate performance versus that of the corporation's industry.
Accessible information is information that can be obtained when needed. Accessibility depends on the customer and even the specific circumstances for that customer. For information quality to occur, timeliness and accessibility should complement each other. Timely information that is inaccessible or accessible information that is obsolete, cannot satisfy an information customer's needs.
Applications of information technology to customer services offer excellent examples of how firms can use quality in information accessibility to obtain competitive advantage witness 800 numbers for computer software and hardware firms, credit card providers and mutual funds.
To paraphrase a famous line, "no information is an island." Information quality lies not only in the quality of the information itself, but also in how it can be combined with other information and delivered to a customer. This often involves systems working together.
The proper information architecture enhances information quality by making the information suitable for enhanced uses. A quality architecture implies a dynamic structure that can grow with changing customer requirements. This is necessary when companies must leverage their information base to develop new products and services as well as optimize their production and management processes.
Historically, information security has been a stepchild of the information technology revolution. Security often was either added on after information systems development was completed or ignored altogether. Two aspects of information security are protecting information from people (logical security) and protecting information from natural disasters (disaster recovery planning). Logical security relies on logical barriers such as passwords, data encryption and transaction authentication, along with human vigilance. Disaster recovery planning involves protecting information and ensuring appropriate back-up and alternate processing procedures are in place.
The topic of information security has become increasingly visible in recent years. Indeed breaches of information security are a potential Achilles heel for the entire information technology revolution because information that is not secure cannot be trusted, and will not be used to its fullest potential.
Information has validity when it can be verified as being true and satisfying appropriate standards related to other dimensions such as accuracy, timeliness, completeness and security. The most common form of verification is auditing information, either as an ongoing practice or as part of a special project. Auditing can uncover mistakes and is a accepted measure of information quality. An obvious example is the corporate financial statement, which to be credible must be audited by an independent party. However, just as quality cannot be inspected into manufactured products, quality cannot be audited into information products.
Validity is a resultant rather than a causal dimension of information quality. Though validity may be high, other crucial dimensions may be low and overall, the information may be of poor quality. Ultimately, optimizing the design and ongoing operation of the human and technological information system is the road to information quality.
Information's value lies not in itself but in how it affects its customer. This effect may be passive, such as the customer's reaffirming or changing a intellectual position, or active such the customer making a decision resulting in an action. Bits of disjoint information are like the brush strokes of a painting which, however well formed, achieve their real meaning only when viewed by an observer. What is important is how the information's customer is affected by the information's picture. This frame of reference is significant because in the age of information technology we tend to think of information as an end, rather than a means. It isn't.
Because the information customer is central to this process, those involved with the design of information systems and the subsequent provision of information should understand the meaning of information quality for their customers. To do this involves:
o Know Information Customers and Suppliers
Knowing the customers and their business needs is a precursor to understanding how those customers define information quality. Businesses must also know their customers if they are to take customer route to improving information quality by improving their customer's information processing capabilities. Information suppliers are important because the information they supply affects what can be delivered. In the short term this may be a constraint, such as being unable to deliver unit product costs because the production shop does not collect data on matching product volumes. In the long term this means how information suppliers can change their information systems to satisfy customer needs.
o Understand the Meaning of Information Quality
Each customer has a different view of the dimensions of information quality. As these views evolve over time, several key questions must be answered over and over again: First, are yesterday's perceptions of quality needs still valid? Five years ago, accuracy and completeness may have been the key factors for an inventory management system. Now, timeliness and format may also be critical. Second, how do quality needs translate into technology requirements? The constant stream of new products and new technologies requires framing the technology implementation decision to ensure fundamental enhancements are separated from bells and whistles. Third, do internal information collection, dissemination and verification procedures measure up to quality requirements? These procedures often were implemented in different customer and technological environments than exist today, and may now be too little or too much.
o Evaluate Current Progress
Progress can be measured by monitoring internal processes and customer perceptions. Some dimensions, such as accuracy and timeliness, can be monitored using standard techniques such as control charts. Benchmarking can also be used to augment internal measurements as well as for monitoring qualitative dimensions such as format and architecture. Because customers constantly redefine information quality, constantly reevaluating products positions along the relevant quality dimensions is particularly important for companies in the information product business.
o Use Technology Rather Than Be Used By It
In today's environment, information quality as defined by almost every dimension mentioned above requires information technology. However, technology can be a double edged sword: use properly and live by it; use improperly and die by it. Organizations who have not defined their information quality needs and who do not understand how a new product affects their fundamental interests, can be led astray. Being unaware of information quality needs may lead either to wasting millions on unnecessary technology or to passing up a breakthrough technology and with it, an opportunity to obtain a competitive advantage by closing a quality gap or redefining a production processes.
(Figure 2 Here)
How the dimensions of information quality translate into business strategy depends on how the information is being used. For companies offering information-based products, the dimensions of information quality can be used to compete in ways similar to how manufacturers compete along the dimensions of product quality. This may mean creating market niches based on several dimensions where a company may see an opportunity such as timeliness, format and completeness; protecting leadership positions along information quality dimensions where the company is already strong; and even de-emphasizing quality along dimensions where customers no longer require it.
(Figure 3 Here)
Creating new niches is an opportunity for many small companies. Established companies can build on existing strengths by strengthening quality dimensions. When information's role is that of component in the production process, organizations should identify the key dimensions of information quality and continuously improve on them. For example, in the apparel industry firms are moving to quick response an answer to reduce inventories and shorten cycle times. Information timeliness is a critical quality component. Many financial services firms are bundling products together. Here, format and compatibility are two key dimensions. Processing business transactions over the internet will become commonplace for many businesses. Here, information security may distinguish winning systems from the losers.
Most organization have significant investments in using information technology for planning and control. Unfortunately, many of these systems were developed years ago and have evolved independent of changing perceptions of information quality. Some were tailored to meet the needs of long-gone executives; others were designed to move information up the organization while today's need may be for information to be disseminated across the organization; others responded to market conditions that are now history. To ensure these information system investments are effective requires answering several questions:
(Figure 4 Here)
o Are yesterday's perceptions of our quality needs still valid? If not, we may be producing the wrong information-based product, missing important opportunities, and may be spending money on unnecessary quality dimensions.
o How do quality needs translate into technology requirements? Technology investments should be linked with what information quality dimensions are important in order to ensure the most effective use of scarce financial and human resources.
o Is our technology strategy consistent with our quality needs? Our current hardware and network architecture and software strategies may be technologically current but they may not support what our customers demand along information quality dimensions.
o Do internal information collection, dissemination and verification procedures measure up to quality requirements? If not, management information flows may lead to an incomplete or erroneous picture of the true situation.
Managing information quality is a continual process. Products and technologies come and go, high quality information today may be low quality information tomorrow. To compete requires a continual focus on definition and follow-through as regards information system design and operational processes. The multiple dimensions of information quality can be used to add structure to this inherent complexity. By defining, assessing, modifying and redefining what information quality is and how it can best be managed, information can remain the vehicle rather than the impediment for achieving business success.
1. Richard B. Chase and David A. Garvin, "The Service Factory", Harvard Business Review, (July-August 1989), pp. 61-69.
2. David A. Garvin, "Competing on the Eight Dimensions of Quality", Harvard Business Review, (November-December 1987), pp. 108-109.
3. David A. Garvin, Managing Quality; (New York: Free Press, 1987).
4. Valarie A. Zeithaml, A Parasuraman and Leonard L. Berry, Delivering Quality Service: Balancing Customer Perceptions and Expectations, (New York: Free Press, 1990).