Data collection methods and tools

THEORETICAL POSITIONING

• Thinking about the question “what is the research process and how am I influencing it?” (Jamieson, et al. 2023), I was considering that I am the entirety of the research process: both participant and researcher. So unavoidable bias and influence will occur, but I will try and minimise these things by not deciding when the study will take place, and doing it retrospectively rather than pro-actively.
• I am also aware that I am unlikely to find anything of any great CO2e impact, about my daily activities in the university, as using devices is as Berners-Lee (2020) suggests a pretty low carbon activity — so my data usage will be, quite minor in terms of its impact. Significantly less than getting a return flight from London to Hong Kong (which I am doing to see my partners family in Feb 2025, this leaves me with bittersweet feelings in terms of its GHG impact). Being cognisant doesn’t allude to or offer me any real thing other than simply being aware of my constant micro impacts.
• As I’m collecting data on myself, critique of auto- data collection is necessary. Alvesson (2003) suggests autoethnography is much too inward looking, and Delamont (2009) harshly rejects autoethnography for various reasons suggesting laziness and introspection, and states it might even represent a worldview of those in power, such as researchers rather than their subjects (Poerwandari, 2021).
• Some however, see small merits, such as Eriksson (2010) who after indicating that autoethnography is in essence, narcissistic — a position which I must say I agree with — goes on to suggest that a method of self-examination could present a better understanding for the interest of others (Poerwandari, 2021).
• This position is furthered by Hamdan (2012) who suggests that auto(ethnographic) researchers might possess insider knowledge which mean they experience the phenomenon from a position of understanding and therefore might be able to better analyse their findings from this position of prior knowledge (Poerwandari, 2021).
• In the defence of my rationale for choosing (auto)self-analysis is predominantly because gathering intimate data about emails sent and received, files uploaded, stored, downloaded, is as current data protection and privacy practices suggest, information that is protected. With a wider scope, sample, and longer timeframe, this data could of course be collected through ethical means and frameworks, but was out of scope for this Action Research Project.
• It may turn out that the methods I undertook could become some sort of teaching or dissemination skeleton framework [AN ACTION?] in order to broaden the data sample and be able to draw comparisons and comparative analysis. It would also ethically be easier for people to gather data on themselves following my protocol of collection.

.

METHODS AND TOOLS

• For my data collection I used a retrospective auto-quantitative direct-measurement method, using a strucrured (almost systematic) approach to the collection and analysis of the data (Sage, 2024). Almost statistical.
• As I can’t seem to locate the correct specific academic terminology on the exact approach I took, we can call the research design in a tokenistic way, ‘innovative’ (Hara, 2015). 
I unpack my use of the above terminology below:
  • Retrospective, as I collected data after the event had happened, rather than try and actively record everything I was doing as I was doing it (to attempt to reduce participant bias through conscious awareness of each of my actions) (Sage, 2024)
  • Auto-quantitative, meaning the gathering of quant data on myself (Sage, 2024)
  • Direct-measurement, by being able to quantitatively look at data on myself (emails sent, received, files uploaded, downloaded, stored etc) using direct collection methods, such as quantitative counting of the information in the sample (volume of emails received) in understandable metrics allows the study to be easily repeatable or teachable (INTRAC, 2017)
  • Structured (almost systematic) I used repeatable, quasi-statistical methods of recording the data by using a spreadsheet and a tool for documentation, and used the same approach repeatedly (Sage, 2024). I admit, that although its rigorous, its probably not actually ‘systematic’ in the eyes of actual research.
• My data collection method, laid out in detail below, was in short: for each of the 10 days of data collection, record in a spreadsheet each email received or sent; mark down if there were others in the email chain, or students if they were in the Moodle ‘quick-mail’.
• I then did an overall data audit to see what, if any files were uploaded or downloaded to OneDrive, and total volume of stored data

• In a way, this data collection sits in the sort of quantified-self / auto-analysis, which has over recent years become popular outside of research circles. This practice is encouraged by the increased availability of devices, tools, and desires to self-analyse and know more about oneself (James Wilson, 2012; Wolfram, 2012)
• However, some argue of negative consequences that occur from quantified-self-analysis, offering us intriguing new framings and nomenclature, such as ‘data-festishism’(Sharon and Zandbergen, 2017) where people are enamoured by the data gathering purely in an autotelic way, or negative overwhelming feelings occur through the auto-gathering of data on themselves, being called an ‘infodemic’ (WHO, 2022).

.

SAMPLE PERIOD AND RATIONALE

• To try and reduce participant bias and measurement bias, I did not specify the 10 days of data collection ahead of time — I decided a range of 4 weeks, working backwards from the end of term (Friday 13 December 2024). Meaning the potential range period of the data collection was between 15 Nov 2024 — 13 Dec 2024.
• The exact start date for the 10 day sample window was decided by a ‘truly’ random number generator (Haahr, 2024), see Figure 1 below.
• I chose 10 days, because its 1/3 of an average month (The mean month-length of the Gregorian calendar is 30.436875 days)

.

DATA COLLECTION METHODOLOGY

STAGE 1

1 — Filter the email box for date range, in this example: 6 Dec 2024

2 — Go sequentially through this list of emails, and if the email chain is obvious, mark down recipients in spreadsheet (see STAGE 5). However, if its from Moodle, as per this example, further deduction is needed to work out roughly who this email was sent to. I then clicked the Moodle link [No 2] to follow to Moodle to find out who it was sent to.

.

STAGE 2

1 — We can now see this is FDD (Fashion Design and Development), but we don’t know which year this Moodle email was sent to.

2 — So, I find the Assessment Brief of the particular unit, and follow highlighted link to see which year the email was sent to.

.

STAGE 3

1 — I confirm the course

2 — I now know its Stage 2-Level 5, which means its BA Y2

.

STAGE 4

1 — I go on to UAL ActiveDashboards > Enrolments Summary Charts  (UAL, 2025)

2 — Select LCF

3 — School of Design and Technology

4 — BA Fashion Design and Development

5 — locate year 2

6+7 — locate the Home and Overseas student numbers for 24/25 academic years

8 — make a sum of home (50) and overseas (39) = 89 students in y2 BA FDD at LCF

.

STAGE 5

1 — sequentially add data to the cumulative spreadsheet, inputting minimal data, and anonymising it where necessary

2 — +89 (as per total from STAGE 4) to column E: Students/others in chain;
+1 to column F: ‘+1 for me’ to be able to add the total for this column to know how many emails I personally received;
+1 to column H: ‘Moodle’, to mark which emails were from Moodle, and which ones were not.

.

---

References