1 Introduction
2 Research Background
2.1 Taxonomy Development in Information Systems and Adjacent Fields
2.2 Taxonomies as Design Science Research Artefacts
3 Research Method
3.1 Status Quo: Data Collection and Coding
Number of articles | (1) Citation analysis of Nickerson et al. (2013) | (2) Keyword search in AIS Senior Scholar's Basket of Journals + BISE | (3) Keyword search in conference proceedings (ICIS, ECIS, PACIS, AMCIS, DESRIST) | Total unique articles* | |
---|---|---|---|---|---|
Phase 1: Identification of relevant articles | |||||
...in the initial result set | 398 | 20 | 71 | – | – |
...after removing articles that are non-English and that neither build nor evaluate a taxonomy | 135 | 17 | 65 | 164 | 100% |
Phase 2: Analysis of relevant articles | |||||
...in which researchers build a taxonomy | 133 | 15 | 64 | 160 | 98% |
...in which researchers evaluate a taxonomy | 51 | 6 | 19 | 56 | 34% |
Number of articles… | (1) Citation analysis of Nickerson et al. (2013) | (2) Keyword search in AIS Senior Scholar's Basket of Journals + BISE | (3) Keyword search in conference proceedings (ICIS, ECIS, PACIS, AMCIS, DESRIST) | Total unique articles* | ||
---|---|---|---|---|---|---|
Analysis of taxonomy building | ||||||
In which researchers build a taxonomy… | 133 | 15 | 64 | 160 | 100% | |
…using the method proposed by Nickerson et al. (2013) | 122 | 4 | 45 | 123 | 77% | |
…using a different or no explicit method at all | 11 | 11 | 19 | 37 | 23% | |
Reporting of the taxonomy building process | ||||||
dedicated method section | 92 | 8 | 41 | 99 | 62% | |
dedicated method reference | 125 | 6 | 49 | 129 | 81% | |
clear development approach | 108 | 15 | 45 | 126 | 79% | |
clear meta-characteristic | 84 | 4 | 36 | 85 | 53% | |
transparent number of iterations | 63 | 4 | 26 | 64 | 40% | |
transparent number of examined objects | 82 | 15 | 36 | 100 | 63% | |
transparent ending conditions | 76 | 4 | 30 | 77 | 48% | |
…reporting the research process transparently (i.e., fulfilling all seven attributes above) | 34 | 4 | 13 | 35 | 22% | |
Reporting of the taxonomy building product | ||||||
clear number of dimensions | 101 | 12 | 55 | 122 | 76% | |
clear number of characteristics | 95 | 11 | 52 | 113 | 71% | |
clear if mutually exclusive | 68 | 6 | 35 | 76 | 48% | |
clear if collectively exhaustive | 53 | 6 | 28 | 59 | 37% | |
…reporting the research product transparently (i.e., fulfilling all four attributes above) | 46 | 6 | 25 | 52 | 33% | |
Analysis of taxonomy evaluation | ||||||
In which researchers evaluate a taxonomy… | 51 | 6 | 19 | 56 | 100% | |
…and report at least one evaluation method | 50 | 6 | 19 | 55 | 98% | |
…and report at least one evaluation criterion | 46 | 4 | 16 | 49 | 88% |
3.2 Extended Taxonomy Design Process and Design Recommendations
4 Status Quo: Taxonomy Design in IS
5 Results
DSR Activities | ETDP Step(s) | Taxonomy Design Recommendations | Example References | |
---|---|---|---|---|
I) Identify problem and motivate | 1–3 | Specify the phenomenon under consideration and justify why a(nother) taxonomy is the right approach to its conceptualization. | ||
1–3 | Specify the taxonomy’s purpose(s) (e.g., describe how and in which context the taxonomy is intended to be used). | |||
1–3 | Specify the taxonomy’s target user group(s) and reflect on how researchers (in a specific research community) and/or practitioners (in a given industry, with specific roles and responsibilities) may benefit from the taxonomy. | |||
II) Define objectives of a solution | 4–5 | Determine a meta-characteristic that specifies the taxonomy’s angle on the phenomenon under consideration. | ||
4–5 | Review and potentially refine the meta-characteristic after the first iteration(s). | |||
4–5 | Determine and justify ending conditions and anticipate evaluation goal(s), in light of the taxonomy’s purpose(s) and target user group(s). | |||
III) Design and development | Choose approach | 6–10 | Start with a conceptual-to-empirical iteration if the existing knowledge base holds relevant insights about the phenomenon under consideration; start with an empirical-to-conceptual iteration if a significant number of objects are available representing the phenomenon under consideration. | |
6–10 | Prioritize conceptual-to-empirical approaches if the taxonomy purpose is normative (i.e., what should be); prioritize empirical-to-conceptual approaches if the taxonomy purpose is descriptive (i.e., what is). | |||
6–10 | Conduct at least one empirical-to-conceptual iteration and at least one conceptual-to-empirical iteration. | |||
Empirical-to-conceptual iteration | 7e–9e | Consider multiple sources for identifying objects (e.g., literature reviews, search engines, databases, and interviews) and justify the sampling strategy (e.g., random, systematic, full). | ||
7e–9e | Consider qualitative (e.g., open coding, axial coding, selective coding, card sorting) and quantitative techniques (e.g., cluster analysis, text mining) to identify characteristics and group them into dimensions. | |||
7e–9e | Consider a second-level grouping of dimensions (e.g., dimensions and subdimensions for organizing characteristics). | |||
III) Design and develop-ment (continued) | Conceptual-to-empirical iteration | 7c–8c | Consider multiple sources for conceptualizing characteristics and dimensions (e.g., literature review, experience, judgement). | |
7c–8c | Account for and refer to existing taxonomies that may inform the taxonomy building. | |||
7c–8c | (Re-)examine objects to validate the new characteristics and dimensions that you have conceptualized. | |||
Create/revise taxonomy | 10 | Use taxonomy operations such as adding, updating (e.g., renaming, swapping, splitting, merging, promoting, and demoting), and deleting characteristics and dimensions. | ||
IV) Demonstration | 11-12 | State which objective ending conditions were met in each iteration or why objective ending conditions were not or only partially met. | ||
11–12 | Ensure that the taxonomy’s characteristics are mutually exclusive and collectively exhaustive; if not, justify it for each dimension. | |||
11–12 | Only accept and clearly state characteristics under which no object is classified if the characteristics result from a conceptual-to-empirical iteration (e.g., when characteristics are expected but not yet implemented). | |||
V) Evaluation | 13–17 | Configure the evaluation in light of the taxonomy's evaluation goal(s), purpose(s) of use and the target user group(s) by determining a suitable evaluation method, at least one evaluation criterion, and a corresponding target value/threshold (answering the ‘why’, ‘how’, and ‘what’ of evaluation). | ||
13–17 | Whenever possible, use new objects for ex ante taxonomy evaluation; only reuse objects from the taxonomy building if few objects are available representing the phenomenon under consideration. | |||
13–17 | Examine whether objects have evolved since the taxonomy building when re-using them for taxonomy evaluation. | |||
13–17 | Recruit evaluation partners (e.g., participants in interviews, focus groups, experiments, case studies) other than those already involved in the taxonomy building process. | |||
VI) Communication | 18 | Document the taxonomy evolution including approach and changes for each iteration. | ||
18 | Visualize the taxonomy (e.g., table, textual, visual) in a way that fits its purpose(s) and target user group(s). | |||
18 | Provide descriptions for each characteristic and dimension. |
5.1 Identify Problem and Motivate
5.2 Define Objectives of a Solution
Evaluation goals | Taxonomy users aim to use the taxonomy to… | Taxonomy designers provide… |
---|---|---|
Describing | … describe a certain phenomenon | … characteristics and dimensions that serve as a basis to describe a certain phenomenon |
Identifying | … identify one specific object that represents a certain phenomenon | … characteristics and dimensions that serve as search criteria to identify one particular object |
Classifying | … classify objects that represent a certain phenomenon | … characteristics and dimensions that serve as scheme to classify one particular object |
Analyzing | … analyze objects that represent a certain phenomenon | … characteristics and dimensions that serve as a basis to determine similarities and differences of objects |
Clustering | … cluster objects based on similarities and differences to consider types of objects rather than individual objects | … characteristics and dimensions that serve as a basis for grouping a set of objects in such a way that objects in the same group are more similar to each other than to those in other groups |
5.3 Design and Development
Taxonomy operations on taxonomy elements (i.e., characteristics, dimensions, subdimensions) | Taxonomy element before taxonomy operation | Taxonomy element after taxonomy operation | |
---|---|---|---|
Add (insert a new element) | |||
Update | Rename (change the name of an element) | ||
Swap (change the order of two elements) | |||
Split (divide an element into at least two elements) | |||
Merge (join at least two elements into one element) | |||
Promote (move an element to a higher level of abstraction) | |||
Demote (move an element to a lower level of abstraction) | |||
Delete (remove an existing element) |
5.4 Demonstration
5.5 Evaluation
Why1 | How2 | What3 | |||
---|---|---|---|---|---|
…concerning function | …concerning environment | …concerning timing | …concerning method | … concerning criteria | |
(11) Check objective ending conditions | Formative | Artificial | Ex ante | Consensus among the taxonomy designers | see objective ending conditions of Nickerson et al. (2013) |
Is it a taxonomy? During the building of a taxonomy, taxonomy designers objectively demonstrate whether the present version of a taxonomy fulfils the necessary condition of being one | |||||
(13) Check subjective ending conditions | Formative | Artificial | Ex ante | Consensus among the taxonomy designers | see subjective ending conditions of Nickerson et al. (2013) |
Is it an applicable taxonomy? During the building of a taxonomy, taxonomy designers evaluate based on their subjective perception whether the present version of a taxonomy fulfils the sufficient condition to be an applicable taxonomy | |||||
(15) Configure evaluation | Summative | Artificial and/or naturalistic | Ex post | see Appendix 3 for taxonomy-related evaluation methods | see Appendix 3 for taxonomy-related evaluation criteria |
Is it a useful taxonomy? After the building of a taxonomy, taxonomy designers evaluate based on the feedback of (potential) users whether the completed version of a taxonomy fulfils the sufficient condition and evaluation criteria to be a useful taxonomy |