Keywords

1 Introduction

This paper shares findings arising from the initial coding iteration of self-reported data from FutureLearn (MOOC) learners, to investigate the participants’ Self-Directed Learning (SDL) experiences. The Bouchard framework [1] presents SDL dimensions using a specific terminology: algorithmic, conative, semiotic and economic. This allows SDL experiences to be categorized and interpreted from four important online learning angles: the pedagogical, psychological, infrastructural, and economic elements. As contemporary online learning is becoming increasingly learner-centered [2–4] it is becoming an increasingly important educational concept.

There is currently a research gap in understanding the full range of SDL dimensions that are used by the learner when s/he engages in an online course [4, 5].

2 Self-directed Learning

In this study, SDL relates to research into adult learning, based on the andragogy concept of Knowles [6], but also embedding technology as an influencing factor for SDL. Knowles [6] described SDL broadly as “a process in which individuals take the initiative, with or without the help of others, to diagnose their learning needs, formulate learning goals, identify resources for learning, select and implement learning strategies, and evaluate learning outcomes’ (p. 18).

Students need to have a high level of self-direction to succeed in mLearning and online learning environments [7]. Learners themselves also consider that achieving the level of self-direction necessary for successful learning in a MOOC is related to prior experience and its resulting self-efficacy [8, 9].

Any SDL framework will need to take into account all these dimensions in a structured way. The Bouchard framework offers such a set of SDL dimensions.

3 The Bouchard Framework

Based on interviews with 40 professionals, Bouchard examined four dimensions of self-directed learning: the algorithmic, conative, semiotic, and economic. Bouchard [1] concluded that only through the careful application of multi-dimensional models can progress be made towards creating environments that truly support the emergence and development of SDL. The work of Bouchard builds upon research performed by Long [10] who provided two fundamental ways in which learning could be learner-controlled: psychological and pedagogical. Bouchard rethought the concepts of Long (psychological became the conative dimension, pedagogical became the algorithmic dimension) and he added two additional dimensions: a semiotic and economic dimension. As a result, Bouchard’s framework [1] seemed well suited to use for a first iteration (bird’s eye analysis), as it provided a schematic for SDL taking into account different dimensions related to online learning.

4 Methodology

To plan and analyse this study, a Grounded Theory (GT) methodology was chosen to organize the different stages of questioning the learners, and to set up research instruments [11]. GT fits research looking for meaning as perceived by the research subjects and permits data like learning experiences to be analysed [12].

4.1 Data Collection

The UK-led MOOC initiative FutureLearn was launched in 2013, and is offering free courses built upon a pedagogy that mixes mobile learning and social learning approaches [13]. This study collected data from three FutureLearn courses, which ran between 1 September and 15 November 2014. The 52 research participants were all experienced online learners. This was a decision based upon outcomes indicating that prior MOOC experience results in more efficient SDL [4, 8]. Experienced online learner covers learners that had prior MOOC and/or online learning experience, and had three or more years of social media experience.

This study uses elicited data (written, digitally delivered, and audio data) from 52 participants who gave full consent prior to the research. The research consisted of three phases to fully capture the SDL experience.

  • Phase 1 – expectations (pre-course): gathering the expectations of the FutureLearn participants by collecting data through an online survey.

  • Phase 2 –learning logs (during course): the participants kept learning logs: filled in every 2 weeks probing the actual learning experiences.

  • Phase 3 – reflections (post-course): structured one-on-one interviews investigating differences between the learning expectations and the actual experiences in regard to SDL.

The data corpus consisted of 792 pages of text coming from learning logs, 115 pages of online survey answers, and 48 pages of interview transcript texts.

4.2 Data Analysis

The data analysis is in its initial coding stage following grounded theory coding suggestions [11], providing first impressions. Constructing theory and relating it to interactions was crucial for selecting GT as a method [11]. In order to analyse and interpret the rich, elicited data, memoing (making researcher assumptions transparent) and the following coding iteration was followed: initial coding, line-by-line coding and focused coding. Allowing the researcher to separate, sort and synthesize large amounts of data. Once the initial categories emerged, those categories were compared to the Bouchard Framework dimensions. The examples below are only a selection of the elements found, due to space restrictions.

5 Sharing Initial Findings

Following the analysis, two of the four dimensions (i.e. semiotic and economic dimension) needed to be revised to fit contemporary, massive online learning (e.g. adding collaborative learning). In addition to that, the algorithmic and conative dimensions harboured elements that needed some updating to match them with current FutureLearn options (e.g., pedagogy, available learner interactions). This resulted in additional dimensions, listed after Bouchard’s known dimensions for each group of elements.

5.1 Algorithmic Dimension

The pedagogical options, and more specifically how the learner uses or interprets them for their own use, are gathered under the algorithmic dimension (Table 1).

Table 1. Algorithmic dimensions.
Full size table

5.2 Conative Dimension

The conative dimension groups elements related to the social and psychological profile, as well as the context the learner is in (Table 2).

Table 2. Conative dimensions.
Full size table

5.3 Semiotic Dimension

In contemporary online learning each type of media possesses its own intrinsic characteristics that facilitate or hinder learning, depending on each individual’s learning preference [1], which brings along a new, wide variety of semiotic dimensions (Table 3).

Table 3. Semiotic dimensions.
Full size table

5.4 Economic Dimension

Bouchard [1] saw the perceived economic value of its knowledge in the marketplace, either as an professional asset or as a means of production (Table 4).

Table 4. Economic dimensions.
Full size table

6 Summary and Future Work

From these preliminary findings promising SDL dimensions were distilled coming from SDL experiences shared by FutureLearn participants. The Bouchard framework needs to be adapted once higher-level dimensions emerge from this on-going study. Once the full analysis is finished, it will potentially reshape Bouchard’s framework, and offer a framework for SDL in FutureLearn courses and MOOCs.