Timelines and time-ordered matrices

Synonyms:
time-oriented matrices, time-oriented networks

Timelines and time-ordered matrices are useful ways of displaying and analysing time-related data.

Creating a timeline or matrix can help to clarify key events and sequences. Having a visual display of qualitative data in one of these forms rather than keeping the information in an extended piece of text can help researchers make connections between events, find or justify explanations, and draw conclusions.

In their book, Qualitative Data Analysis, Miles and Huberman (1994) discuss the use of visual displays in the analysis of time-oriented qualitative data, focusing in particular on two display types: matrices, a grid-like display with defined rows and columns, and networks, a series of linked nodes, of which the timeline is an example.

On the importance of such displays in the act of analysis, Miles and Huberman (1994: 90-1) write:

“Our experience tells us that extended, unreduced text alone is a weak and cumbersome form of display. It is hard on analysts because it is dispersed over many pages and is not easy to see as a whole. It is sequential rather than simultaneous, making it difficult to look at two or three variables at once. It is usually poorly ordered, and it can get very bulky, monotonously overloading. Comparing several extended texts is very difficult. ... Valid analysis requires, and is driven by, displays that are focused enough to permit a viewing of a full data set in the same location, and are arranged systematically to answer the research questions at hand.” (Miles and Huberman 1994, p.90-1)

Examples

An example of an effective timeline used for a report can be seen in the Early Childhood Nutrition and Anemia Prevention Project. This example, drawn from the Summary Report, provides a clear overview of a large amount of qualitative data that aids in the task of processing the information for both the researchers and stakeholders.

In a presentation for AEA Evaluation 2013 conference, this First 5 LA (archived link) timeline is examined in detail by Artineh Samkian, the University of Southern California, and Joel Greene, Harder+Company. As can be seen, the timeline is clearly laid out, with the nodes of information linking to each other logically, allowing connections to be easily seen. This presentation also contains other examples of timelines to give an idea of the range of styles for different purposes and audiences.

In their paper for Nurse Researcher, Williamson and Long (2005) walk readers through the process of creating various types of data displays for qualitative data analysis. There is a focus on how these displays can work together to meet the research's key objectives. The image shown below is part of their example of a time-ordered matrix. The full paper and image can be found at 'Qualitative data analysis using data displays'.

Advice for using this method

Adapted from Miles and Huberman (1994: pp.93-101)
  • Choosing the form
    • Thinking about the format of your data display should optimally begin in the data collection phase, and be driven by your research questions and aims. This way, you can tailor your questions and data collection to the information you will need for your timeline or matrix
    • Time-ordered matrices are useful for:
      • Giving an overview and understanding of an extended process of events
      • Can be useful in developing causal explanations
      • Creating displays of both long and short time frames
    • Timelines, or 'networks' (Miles and Huberman 1994), are useful for:
      • When there's a need to focus on multiple variables at once
      • Holding a large amount of information
    • After choosing a form or creating a rough draft of your timeline or matrix, you should be open to the fact that this initial form may need to change as you collect more data or conduct some preliminary analyses
      • New data may  “qualify, put in perspective, and disqualify earlier ones” (Miles and Huberman 1994, p.98)
      • Having a fixed form may “bully the data into shapes that are superficially comparable across cases, but you actually may be comparing intrinsically different things - on dimensions that turn out to be trivial” (Miles and Huberman 1994, p.98) .
  • Organising your data
    • Code your data effectively to make it easier to organize into your display.
      • If possible, this can useful to do during the data collection phase.
    • It can be useful to also a reference of where the data came from in your field notes so that it can be referred back to easily.
  • Deciding what data to include:
    • There are no set rules about what data to include and exclude, think about your purposes and aims in creating your display.
    • It's recommended to keep a record of the decision making process that goes into choosing which elements to include (Miles and Huberman 1994, p.100).
    • While it's important not to overload your visual display with huge swathes of text, Miles and Huberman (1994, p.100) warn against  "overreducing data[, which] can obscure understanding”.
    • It can be useful to use a system of file cards or similar to help define which events are most important to include.
    • The types of data you use might include the following (Miles, Huberman & Saldaña 2014, p.116):
      • Direct quotes or extracts from field notes
      • Summaries or abstracts
      • Explanations
      • Judgements or ratings about effects
  • Making your display
    • Choose time periods that are suitable to your needs, aims, and sequences that you are trying to explore.
    • Play around with design features such as text boxes, colours, and fonts to make key events and connections easier to see.
      • If the display is also going to be used in presenting the data to stakeholders, it's useful also to think about what other design features can be incorporated for visual effect. Photos and graphic images can be powerful additions.
    • Be ready to adjust or redesign your display if you feel this is necessary as more data emerges or after preliminary analysis.
  • Analysing the data
    • Start analysing the information after, or during the creation of, a first draft of the display.
      • This will allow you gain an idea of what information that has been included can be left out, and what information should be added. (Miles and Huberman 1994, p.114)
      • You may also see ways that the organisation of the display needs to be improved. This may be the grouping of different elements together or separating out of events to better understand the sequence.
    • Start by giving the display a quick scan for the most prominent information, then go back over it in more detail.
    • Use the advantages of the display to their fullest by:
      • Looking for sequences, causes, changes in and stability of events.
      • Noting patterns, contrasts and key themes that emerge.
      • Constructing chains of evidence for decision making or conclusions.
    • As you are analysing, write down any conclusions you make. This can add clarity to your thoughts and allow you to see new ideas (Miles, Huberman & Saldaña 2014, p.117).
    • Return to your original data to check your conclusions.

Resources

Guides

Examples

Tools

Aquino D, Marley J V, Senior K, Leonard D, Joshua A, Huddleston A, Ferguson H, Helmer J, Hadgraft N, & Hobson V. Early Childhood Nutrition and Anaemia Prevention Project. [Summary Report]. Darwin: The Fred Hollows Foundation, Indigenous Australia Program; 2013

Miles, B. M. & Huberman, A. M. (1994) Qualitative data analysis: a methods sourcebook. California: SAGE Publications.

Miles, B. M., Huberman, A. M. & Saldaña, J. (2014) 'Chapter five: designing matrix and network displays'. In Qualitative data analysis: an expanded sourcebook (3rd Ed., pp. 107-120). California: SAGE Publications.

Williamson, T. and Long, A. (2005). 'Qualitative data analysis using data displays' in Nurse Researcher, 12(3): 7-19

Samkian, A. and Greene, J. (2013, October 17). Visualizing Process: How to Create a Stakeholder-friendly Graphic Timeline of Process Data. Presented at the American Evaluation Society Evaluation 2013 Conference, Washington.

Samkian, A. and Greene, J. (2014, January 14). Artineh Samkian and Joelle Greene on Graphic Timelines to Capture Qualitative Process Data [Web log post]. Retrieved from http://aea365.org/blog/artineh-samkian-and-joelle-greene-on-graphic-time...

'Timelines and time-ordered matrices' is referenced in: