Indi Young: Mental Models – review

  • By Régeni Anna
  • 3 Apr 2018
  • No comment
  • 12 mins

About Indi Young
Indi Young started her career as a software engineer, with a Computer Science B.S. graduating in 1987, in her early projects, she realized that there was a gap between what computer scientists knew about the users, as engineers and creators, and what people were trying to ultimately accomplish. This was the recognition that turned Indi Young towards UX research and led to the foundation of UX agency Adaptive Path in 2001.

About the book
Indi Young’s book is considered unique in its class since the scholarly research methodology and the philosophy underlying UX is intertwined in it than usual, plus it extends well beyond the general UX researcher approach. While the objective of UX research is indeed to bridge the gap between the ideas of users and those of the industry about a given product, Indi Young’s book stresses that this area is about a lot more than just that. It is not the product, but the user experience which is in focus: the user’s daily routine, and their ideas and preliminary perception of the world. If we are willing to devote more energy to getting to know the user, we can gain more valuable insights than if we were to get to know them strictly from a product perspective.

By identifying the mental models, not only the product can be designed more intuitive, but using gap analysis (more on this later), we can possibly also perform more developments, some that are truly warranted. Gap analysis is an operation that identifies existing needs.

Chart 1: Indi Young’s mental model diagram.

Chart 1: Indi Young’s mental model diagram. Source: https://blogs.msdn.microsoft.com/canux/2008/02/05/podcast-chat-with-indi-young-on-mental-models/

The part below the horizontal line groups the existing functions of the webpage /service under the relevant towers of mental spaces.

Content Maps. These maps summarise the services of the webpage.

Content Maps. These maps summarise the services of the webpage.

The grouping can be by various types, but the classification as envisaged by the designer will always prevail. Once the mental model and the content map are ready, one can start grouping the function corresponding to the need under the right tower of the mental model. The “gap” featured in the name of the operation refers to the fact that the gaps remaining after the grouping may carry relevant information. Namely, it is based on these gaps that we can identify which functionalities that users would need are still missing as well as any unnecessary functions featured in the content map that cannot be classified in our mental model.

The concept of mental model
First of all, let me quote the definition of mental model taken from the book: “Mental models give you a deep understanding of people’s motivations and thought-processes, along with the emotional and philosophical landscape in which they operate.” Mental models offer deep knowledge of the motives and thought-processes of people; and along with these, the emotional and philosophical framework in which they act in their everyday lives as well.

The greatness of mental models is that they capture the typical behaviour of a person. They give quasi conditioned answer not to a certain situation or issue, but on the contrary: Indi Young specifically points out that from the perspective of mental models, the tools with which we perform a task is  irrelevant. The applicability of mental models covers a spectrum considerably broader than for example a usability test. The mental model describes a typical behaviour which is tool-independent.

According to author of these lines, the best metaphor of the book for the mental model is language. In this metaphor, the model itself is the language, while grammar rules are the links between the tasks. Similarly to the constructive nature of language, the mental model also enables us to express just about anything. Language is not merely the sum of words; it is a system offering an infinite combination of possibilities, which means that it is applicable not only to a single issue, but universally and in a broad spectrum.

Why and what is the mental model good for?
… and is it worth the time, the money and the effort invested? The pages of the book reveal a structure which explains its own raison d’etre with its simple existence. What points beyond its universal applicability, already mentioned as one of its core benefits, is that, first, it remains valid long-term since people’s habits and daily routines do not change overnight, and second, the model is able to give predictions about the new functions that are really needed. This enables us to replace or substitute the intuition and luck factors, so important in design, if we are not surrounded by another Steve Jobs. This predictive analysis performed using the mental model is called ‘gap analysis’, as mentioned above.

The other huge advantage of mental modelling is that we can map from it the product’s information architecture [IA – information architecture]. This is where we can best see how valuable data collection is and how data can help us construct a diverse, multilevel architecture; what is more, the beauty of the method is that a mental model practically builds itself.

Yet another benefit of building mental models is that they bring us closer to the concerns and the daily routines of our users/potential users, help us understand the logic of their thoughts connections, and in return, make us more empathetic towards them. We learn how to see with the eyes and how to think with the head of the user. And empathy is the root of UX; without it, there is nothing to talk about.

The nature and the steps of mental model research
Mental model research belongs to the group of qualitative research methods. As opposed to its peer (quantitative data collection), the essence of this method is not to collect the greatest quantity of quantifiable data using questions specified in advance and then analyse them using quantitative statistical methods, but rather to discover the opinion and habits of the research subject using open questions. With qualitative analysis we cannot quantify our data (or this quantification is difficult), for instance in the case of an online questionnaire-based research, but we are able to answer the causes (why) and see the thoughts and the reasoning of our subject with the eyes of an insider.

In addition, from the preference/evaluative/generative trinity, mental model research belongs to the latter one as it points beyond the evaluative research in the sense that instead of examining the use of a given tool, it takes a step back and observes the general steps the user considers taking in order to reach the goal, while ignoring the tool with which such goal is (or is not) attained.

The researcher conducting a mental model research interview uses open-ended questions and is actively listening. The collected data are recorded in the form of transcripts and voice/video recordings, then analysis starts by “combing” them as part of which they pinpoint the “atomic tasks”. These atomic tasks form the cornerstones indispensable for constructing the mental model and from which the forms and the patterns characterizing the mental model are created.

The research process of mental models as understood by Indi Young
In the following part we share the steps of constructing a mental model as described in the book to enable the interested reader to gain a somewhat more accurate picture about the process of mental model research.

I. Selecting the target group
1. Starts as part of a brainstorming session involving several staff members, where the objective is to define the possible tasks that come up in a given situation. For example, what tasks may turn up on the occasion of a visit to a movie theatre? We invite our friends, we pick a movie, we agree on the time, we check the language of the screening, we search for the movie theatre where the movie is screened in the original language, we agree on the place to meet, we buy some snack for the movie, then afterwards we pick a place where we can discuss the film or just have a chat because we haven’t seen each other for a while….. and so on. Usually a team is able to collect between 150 to 200 tasks on the occasion of such brainstorming, and then narrow them down to around 75 by merging similar tasks. Once this is done, the next step can come, i.e., grouping the tasks according to their behaviour affinity.

2. Grouping the tasks according to their behaviour affinity
This step actually focuses on the persons executing the tasks, since the grouping is based on the identification of the tasks that arise simultaneously. Naming the persons executing the tasks facilitates the resolution of this exercise. One task is not exclusive, there may be several persons executing it. The tasks grouped under the various executors and the similarities among the executors help reach the third step:

3. Creating and naming the audience segments, or a bit misleadingly, the target groups
Creating these audiences is done based on task patterns and the executors. If we manage to obtain clearly differentiated executor groups right away, then we are lucky. But in general, many tasks cannot be clearly linked to one type of executor. In such case it makes more sense to create a smaller matrix where we list the tasks and our possible audiences. For every audience we mark the tasks they have possibly executed, then we proceed with the patterns obtained. On the occasion of yet another brainstorming session we put the sufficiently similar audiences next to one another, then we form our audience segments from the obtained task and executor groups. The next step involves the definition of the audience segments most relevant for the business strategy. This step also has a method developed by Indi Young, but we will not cover this in detail in this blog.

Source: https://www.theodysseyonline.com/if-the-zodiac-signs-were-dogs

Source: https://www.theodysseyonline.com/if-the-zodiac-signs-were-dogs

4. At first glance, the dogs featured on the above picture are all different. But if we take a better look at them, we can group them based on different characteristics (length of fur, body structure, size, fur colour, gender, age, temper… etc.). We can group people in the same way based on the factors relevant for the research.

The practice of recruiting and selecting the participants matching the audience segments is also a key part of the methodology as it does make a difference to what extent our subject really or only theoretically meets the recruitment criteria. For example, in order to avoid any disappointment, it is highly recommended to preliminary test whether our subject is able to and willing to meet such criteria. The book contains a detailed description for this part which helps to select the adequate subjects and to minimise the disappointment that our interview subject is not adequate (e.g., he/she does not want to or is unable to talk for one hour, they only come for the reward, but do not at all meet the criteria….etc.).

II. The survey
1. After the careful selection of our research subjects, the next important step is to define our research framework before doing anything else.
The first step is to once again review what we want to obtain from our research. What are the issues and what are the good solutions that we see; what are the business objectives stated by the stakeholders of the business, the attainment of which they expect from the research and the implementation of its results.

2. Then we conduct interviews with the stakeholders of the business and make sure that they are also aligned and everyone is on the same page in terms of the objectives. To map existing issues, research results available from previous surveys may be excellent sources (focus group results, questionnaires measuring customer satisfaction, complaints).

3. Once we are done with the above-mentioned matching and the exploration of the “field”, it is important to be clear about the degree of details we want to achieve in our research. For example, if you want to map the daily tasks of the cleaners, you must pay attention to what extent the information obtained is useful and from what point on it is unnecessary. Do we care about the order in which they vacuum, wash the floor or do the dishes, or do we only care to know that they execute these tasks. Or is it important for us to know which detergents they use and how much of them they use, or is this information irrelevant?

4. After giving the objective of our research and the degree of details some thought and after having consulted with the stakeholders and selected the appropriate subjects, the next step is to conduct the interviews and prepare the transcripts. Indi has collected a couple of useful advices for perfect data collection and interviewing and she shares them with the readers to enable them to deliver quality work for their own mental model research.

III. Assessment of the data
The assessment of the word-by-word transcript of the interviews is a time-consuming job requiring a high level of attention and is composed of several parts.

1. The first step is to highlight/comb the behaviours, or as Indi calls it, the tasks in the interviews. Defining the tasks is basically the most important part of the entire method as these tasks will provide the skeleton of the entire method, the entire model is constructed out of them.

In Indi’s interpretation, everything that is “an action/operation, idea, feeling, philosophy and motivation, everything that leads the individual to implement an objective” is a task. But this is still not enough to understand what Indi Young really means by “tasks”, but this is a good starting point. Maybe the easiest way to understand what differentiates a task from a “non-task” is if we keep in mind that in the transcript we must identify and not figure out the tasks.

That is, we try to exclude our predicting nature and we only consider a task that is said to be one, and whatever is only a desire or an expectation is not considered a task. We can save those for later as these data are not useless, but we do not take them into account for the list of tasks. We differentiate several types of tasks such as the simple task, the implicit task, the task performed by a third person, the philosophy and the feeling.

Fortunately Indi also helps us identify where we need to be suspicious (already during the interview) as a task may be hiding there and we just need to expose it. The following suggest some tasks to be exposed: medium, statement of facts, explanation, circumstance, complaint. Example for these: If the task itself is that I have to cook lunch; then the medium is the stove, the oven and the pot and the bike I use to ride to the store; the statement of facts is that we have a composter where I can throw organic waste; the explanation is the way I need to properly clean a carrot; the circumstance is that I have little time for cooking; and the complaint is that the store where I can get the missing ingredient is far away. (It is good if this part already starts during the interviews so as not to break the dynamics of the analysis.)

Once we have collected the tasks in an Excel file or on post-its, we can start grouping them.

1. We start grouping the tasks from bottom to top, by combining the related and similar atomic tasks, and creating a broader set, i.e., the tasks.
2. By matching the tasks, we create yet another broader set, the so-called task towers.
3. The top level in this hierarchy is the level of mental spaces, which gives a kind of insight into our users, their goals and approaches as well as the philosophies belonging to the given mental spaces.
4. Once we are finished constructing this hierarchy, all we have to do is to visualize our model.

The book provides many pointers and practical advice for every one of the above-mentioned steps enabling any researcher who has never done such research so far to prepare their first mental model. The book will help the reader properly identify the atomic tasks, group them appropriately and name the task towers and the mental spaces. Moreover, the book will also guide us in selecting the right tools for preparing the mental model, and Indi Young even supplies a Python script which is able to handle any Microsoft or Excel file as input and the output will be the diagram of the mental model.

Overall it is important to highlight that Indi Young’s book titled Mental Models is a handbook with many many (sometimes unnecessarily too many) details, professional advice and exercises for every possible situation of mental model research. It provides guidance for preparing, running and evaluating such research; but you should know that it is not entertaining literature.

Whatever the case, she definitely managed to convince me that mental model research is not only a new buzzword in the field of UX, but a well-thought-out method that may largely contribute to turning user experience into something really full of experience.


ABOUT THE AUTHOR

She has obtained a degree in Cognitive Science from Budapest University of Technology and Economics. She is currently interested in the fields of psycholinguistics, statistics and user experience research & design. Within UX research, her attention is captured by designing and evaluating user tests.

LEAVE A COMMENT

 
Please enter your name
Please enter your comment