Developing user orientated Information Services

The 21st century has been described as the “information age”. This information can is stored and can be retrieved from various sources such as books, journals, databases, CD-ROMs and the internet. However due to the large amounts of information available, users seeking information are faced with the problems of poor knowledge of hardware, user-unfriendly interfaces, difficulties in formulating queries, little knowledge of requests and practices, etc. As a result, various models have been created that attempt to identify how to create information retrieval products and services and what factors need to be taken into account in relation to the user when constructing such products and services. However these models vary a great deal and whilst some of these models concentrate on studying a group/community in general and their information seeking needs, other models are less broad and concentrate on the actual moment of interaction. This report will be attempting to evaluate, compare and contrast and assess the strengths and weaknesses of three these models. The report will also be attempting to assess the usefulness of these models in helping one to study a community.

To conduct my evaluation, I will be using the two models of Wilson (1996) and Spink’s model of Information seeking.

Wilson’s 1996 model of information behaviour

Model Description

Wilson’s model of information behaviour is designed to study group/community and user studies/information behaviour in general. The model is therefore designed so that it can be applied to any such community. Wilson’s model is based on 3 key elements:

information need and its drivers, i.e., the factors that give rise to an individual's perception of need

the factors that affect the individual's response to the perception of need

the processes or actions involved in that response

The model uses a goal-directed “problem solving” approach that concentrates on a “person centred” approach instead of a “system centred” approach and is therefore based on a qualitative method rather then a quantitative method. Wilson states that an individual’s information seeking behaviour is prompted by their physiological, cognitive and effective needs. Wilson, goes on to note that the context of any one of those needs may be the person themselves, or the role demands of the persons work or life or the environment (political, economic, technological etc) within which that life or work takes place. Wilson theorises that the barriers that impede the search of that information will arise out of the same set of contexts.

Wilson’s model is divided into five phases. The activating mechanism phase of the model is characterised by feelings of uncertainty and vagueness in regards to the information need. The intervening variables attempt to assess the psychological, demographic, interpersonal, environmental and source characteristics that may be providing an obstacle for the user. The model also defines the activating mechanism for the information seeking behaviour and outlines that this may be based upon the risk/rewards theory or the social learning theory. Finally, the model attempts to identify the information seeking behaviour of the user which may vary between passive attention, passive search, active search and ongoing search.

Strengths of Model

The model is very general and can therefore theoretically be applied to any information seeking group/community

interactive search episodes provide the informational framework to the problem-solving process through which the user's uncertainty level is reduced

The model takes into account psychological factors such as the thinking processes, analysis and filtering involved on the part of the user.

The model takes into account demographic factors and identifies that there maybe problems associated with the system being used by varying age groups with varying levels of knowledge

The model also considers environmental ...

This is a preview of the whole essay

Strengths of Model

The model is very general and can therefore theoretically be applied to any information seeking group/community

interactive search episodes provide the informational framework to the problem-solving process through which the user's uncertainty level is reduced

The model takes into account psychological factors such as the thinking processes, analysis and filtering involved on the part of the user.

The model takes into account demographic factors and identifies that there maybe problems associated with the system being used by varying age groups with varying levels of knowledge

The model also considers environmental factors i.e.- users with dyslexia and how this may effect the design of the system

The model also identifies that the product or service to which the model is applied may be for business or personal use and therefore this would have a direct effect on the needs of the user, the psychological factors such as the thinking process involved, style of learning etc.

The system is also positive in identifying that users have varying information seeking behaviour that is often dependant upon the individuals needs, psychological, demographic and environmental factors. As a result it takes into account the type of search that the user is conducting and can cater for their information seeking behaviour accordingly.

Weaknesses of Model

The model fails to take into account that different users have different needs i.e. a library database designed for students in general using the Wilson model would ignore the fact that a first year student may have totally contrasting needs to a PHD student.

Although the model takes psychological factors into account, it is unclear whether it takes into account the fact that different users have varying levels of knowledge and therefore their comfort levels, needs and requirements will all vary as a result of that.

Although the model incorporates the “stress/coping theory” as part of the activating mechanism, it fails to take into account that users have varying levels of comfort and levels of precaution when using a new system. As a result, some users may take longer then others to cross that “comfort zone” and may therefore require varying procedures and techniques to help them reach that “comfort level”.

The model fails to take into account the factor of user education and that certain users are more suited to certain methods of education then others. User learning methods are divided between auditory, visual and kinaesthetic and therefore it is important to gain a balance between the user and the learning method as this has a direct effect on barriers such as comfort levels.

The model also fails to take into factors such as the interface design as it will vary according to the product and the interface design such as metaphors, icons, graphics and colour coding etc which would all have a direct impact on how quickly users learn and their levels of comfort with the new product.

The model fails to take into account the “holist” vs. “serialist” approach to learning and the fact that whilst holists prefer to see the “big picture”, the serialists prefer a narrower, broken down approach.

The model offers no method of evaluating the product or service in terms of consistency, coverage, timeliness, accuracy, ease of use, functionality and support/training. As a result, it is very difficult to measure the true effectiveness of the model and to build on any weaknesses.

Conclusion

The Wilson model is a very good general model that takes some important factors into account and can be applied to a wide variety of information products/services. However, the model also has various flaws that are mentioned above that limit its effectiveness and the situations in which it can be applied to.

Spink’s model of Information seeking

Model Description

Spink develops a model that is designed to be most compatible with internet searching and includes feedback as a key element in the information seeking process. Feedback is an interactive process for query modification. Relevance feedback is one of the examples, in which a user selects a small set of items relevant to a query, and the system then uses features from the selected items to adjust the query.

Spink suggests that a search process consists of user judgements, search tactics or moves, interactive feedback loops, and cycles. As a result, unlike the Wilson model, Spink’s model is concerned with the moment of interaction i.e. when people are interacting with information retrieval systems.

Spink illustrates the model as follows: (Source from: Spink,A. (1997) Study of Interactive Feedback during Mediated Information Retrieval)

One or more search strategies constitute an interactive search process

Each strategy may comprise one or more cycles

Each cycle may include one or more interactive feedback loops: user input,

system outcomes, user explanation and judgement and user input

An input may be considered as a move (change) in search strategy or a search

tactic of the advanced search

A user input or query indicates a move

Strengths of Model

The model is designed to be very narrow and concentrated as it is concerned with the moment of interaction i.e. when people are interacting with information retrieval systems. As a result, the model is more focused to a particular information product/service such as internet searching in contrast to the Wilson model which is more general and broad.

Feedback is a major part of the model and therefore it allows the model to be evaluated more easily against criteria such as consistency, coverage, timeliness, accuracy, easy of use, interface design, support etc. This allows the information product/service that this model is applied to; to be evaluated more easily and allows it to incorporate changes and improvements.

The model also takes into account various search strategies which is a clear strength as it is more user friendly through its appeal of offering various search strategies to suit different users. This increases the models user friendliness and functionality.

Weaknesses of Model

The model fails to take into account that different users have different needs i.e. a library database designed for students in general using the Wilson model would ignore the fact that a first year student may have totally contrasting needs to a PHD student.

The model fails to take psychological factors into account and does not consider the fact that users have varying levels of knowledge and therefore their comfort levels, needs and requirements will all vary as a result of that.

The model fails to take into account the factor of user education and that certain users are more suited to certain methods of education then others. User learning methods are divided between auditory, visual and kinaesthetic and therefore it is important to gain a balance between the user and the learning method as this has a direct effect on barriers such as comfort levels.

The model also fails to take into factors such as the interface design as it will vary according to the product and the interface design such as metaphors, icons, graphics and colour coding etc which would all have a direct impact on how quickly users learn and their levels of comfort with the new product.

When applied to a search engine, the model can have key limitations due to language recognition:

cannot handle different words having almost the same meaning (throttle & accelerator, road & street)

cannot handle the same words having different meanings i.e. words like bar (as in soap bar vs. place for refreshment), or at the phrase level, juvenile victims of crime vs. victims of juvenile crime)

cannot handle the same concept represented in a different language, in this case different because of speakers perspective on the concept (unfortunate penalty miss vs. tragic football disaster)

The model is concerned with the moment of interaction and thus is very narrow and concentrated.

Usefulness of models in studying a community

Overall, I feel that both the Wilson model and Spink model can be used to study a community to an extent. Due to the general nature of Wilson’s models compared to the specific nature of Spink’s model, I feel that a mixture of the two models would perhaps offer the best way to study a particular community.

Similarly, Wilson’s model is very good in understanding the psychological, demographic, inter-personal, environmental and source characteristic intervening variables of user communities. If this is applied to a community such as journalists, the model does well to identify their information seeking behaviour through options such as passive search, passive attention, active search and ongoing search. This offers various search mechanisms for the different users with different needs in this community. However a community such as journalists would often require more then a single source of information and thus by incorporating the Spink model to various information products such as databases and CD-ROMS and then incorporating them into a single large database based upon the Wilson model may satisfy the needs of this community.

In contrast, the models fail to take into account that different users in different communities have varying levels of knowledge and thus it may be difficult to apply a single system that would satisfy the needs of all level of users within a community. This can be seen in the case of the student community where a single database such as Dialog can not be expected to appeal to all users it is a database that is difficult to use for the novice user. Similarly, different users within a community may have different needs and this can once again be seen in a student community where a database designed for further reading would have to hold different levels of information for a first year student in comparison to a PHD students. As a result, although it may be a single database serving a single community of users, the needs of these users may be very different.

The models also fail to specify exactly how they understand the needs, wants and demands of the user and determine what the user values. This is vital when assessing what information service to provide to a community and how it will best serve the user through meeting their needs.

The models also fail to take into account user education. Different people learn better using different methods and therefore whilst some users may be more suited to an auditory method such as a presentation, other users may be suited to visual and kinaesthetic methods through a more “hands on” approach to learning. It is vital to incorporate a learning system into these models which identify how the user would gain the information skills/literacy and how best a balance can be gained between the user and the learning method.

Attention and memory constraints suffered by users are another factor that the models do not directly take into account. Metaphors, icons, graphic and colour coding are all factors that can determine how easily the user relates to the product and therefore how quickly they learn.

In conclusion, whilst the Wilson and Spink model both have their own respective strengths and weaknesses, I feel that it would depend upon the particular community itself and its needs that would eventually determine the success of the model. In many communities such as journalism, a combination of the broad Wilson model and a more specific model such as Spink’s model would be very useful as incorporating the two together would help serve the varying needs of the smaller communities within the general community of journalists. I also feel that the models need to incorporate other factors as well such as identifying that different users have varying level of knowledge and expertise, comfort levels, differences in seeking behaviour, tasks goals etc and therefore a single model can not apply to every one in every community.

Finally, I feel that the models need to incorporate a research aspect and attempt to identify how each model can be promoted and taught to the user, taking into account that each user has different information literacy rates and is suited to varying methods of learning. This can also be combined with an assessment aspect of the model which can allow the community to evaluate the product/service that has been based on the model and measure its effectiveness, benefits and performance.

Marwan Nawaz

Bibliography