The Finnish government embarked on an extensive reform of vocational education and training (VET) in 2015. One of the main aims of the reform was to increase work-based learning. The new legislation came into force in January 2018 and consequently it has lead to changes in the steering and regulation system, and introduced a new funding model with a focus on qualifications and employment to improve the effectiveness of VET. However, traditionally Finland has represented the statist skill formation regime combining low involvement of employers in VET and high degree of public commitment (e.g., Busemeyer & Schlicht-Schmälzle, 2014).

Social media can be used by policy makers to support the implementation processes of reforms (Joseph et al., 2017), but in recent years, social media has also proven to be a powerful tool in protesting against educational reforms and policies (Berry & Herrington, 2012). As a whole, examining the social media data may provide a better understanding of how policies are perceived (Joseph et al. 2017). Monitoring social media can be theoretically understood by situating it in theories of public opinion (Anstead & O’Loughlin, 2015). In the context of education policy, public opinion, i.e. individuals’ policy preferences and attitudes, are shaped by (self-)interests and political attitudes, but also by the institutional set-up of the education system (Busemeyer & Garritzmann, 2017). However, it is not always clear whether individuals rather support those kinds of institutions that already exist or policies that are different from the institutional set-up (Busemeyer & Garritzmann, 2017).

This paper seeks to examine the VET reform in Finland by collecting Twitter data and discussions about VET reform. Even if the Twitter data does not represent the whole population, investigating the opinion of the Twitter users may be interesting, since they represent a form of activated public opinion often comprising stakeholders and the ones who are able to promote or contest reforms (Ceron & Negri, 2016). Unlike in many other social networking sites, in Twitter data is publicly shared. Besides published twitter postings or tweets self-provided user information and interaction data can provide additional details about the characteristics of the users (e.g., Kwak et al., 2010). In order to provide information on influencers and public opinion from Twitter data, three research questions are addressed as follows: (1) who are active in Twitter discussions related to VET reform, (2) what kind of networks and communities do these people or stakeholders build, and (3) what are the sentiments of the tweets related to the VET reform? In addition, the longitudinal study during the implementation phase of the reform allows identifying changes in discussions over time.

Social media analytics refers widely to developing and evaluating tools and frameworks to collect, monitor, analyse, summarize and visualize social media data (Zheng et al. 2010). This study employs multiple tools for gathering insights from Twitter data. In the data collection phase, tweets with either of the two hashtags, #amisreformi (VETreform) or #reformintuki (reformsupport), are extracted with two different tools. The first one, TAGS, is a free application to collect tweets. The second on, Luuppi (The Loop), is a machine learning-based tool developed at Turku University of Applied Sciences. Both of these tools allow following social media perceptions both qualitatively and quantitatively. First, to recognize active, visible and engaged users and influencers the ratios of tweets/user and retweets/user are investigated. Following this, the networks are recognized by applying a social network analysis and community detection algorithm to measure modularity (Blondel et al., 2008), which will be visualized with Gephi. To explore the affective dimensions of the expressions, the sentiment analysis will be applied to identify negative, neutral and positive sentiments (Bae & Lee, 2012; Jussila et al. 2017). In order to further explore the experienced challenges or tensions of the reform process, the more precise topics of the tweets expressing negative sentiments will be mapped by a researcher.

The data gathering continues and the final analyses have not yet been carried out. Since the new legislation came into force on 1 January 2018, the number of tweets is 679 for #amisreformi (VETreform) and 124 for #reformintuki (reformsupport). It seems that next to national administration, a versatile, but a relatively small group of actors are active in Twitter discussions. In the field of VET, participants are education providers, vocational teachers and teacher educators, but it is noteworthy that students are not expressing their views or participating in Twitter. So far, the few participating interest groups, such as employee unions, employer unions and student unions, seem to express positive or neutral sentiments towards the reform. Currently, over 70 per cent of the tweets contain positive, around 20 per cent neutral and less than 5 per cent negative sentiments. The negative sentiments at this point relate to diminishing school-based education and challenges with finding workplaces for training. Even though this proposal especially discusses the Finnish reform of VET, the methodologies used may be applied in the wider European context.

Anstead, N., & O’Loughlin, B. (2015). Social media analysis and public opinion: The 2010 UK general election. Journal of Computer-Mediated Communication, 20(2), 204-220. Bae, Y., & Lee, H. (2012). Sentiment analysis of Twitter audiences: Measuring the positive or negative influence of popular Twitterers. Journal of the American Society for Information Science and Technology, 63(12), 2521-2535. Berry, K. S., & Herrington, C. D. (2012). Tensions across federalism, localism, and professional autonomy: Social media and stakeholder response to increased accountability. Educational Policy, 27(2), 390-409. Blondel, V. D., Guillaume, J-L., Lambiotte, R., & Lefebvre, E. (2008). Fast unfolding of communities in large networks. Journal of Statistical Mechanics: Theory and Experiment. doi:10.1088/1742-5468/ 2008/10/P10008 Busemeyer, M. R., & Garritzmann, J. L. (2017). Academic, vocational or general? An analysis of public opinion towards education policies with evidence from a new comparative survey. Journal of European Social Policy, 27(4), 373-386. Busemeyer, M. R., & Schlicht-Schmälzle, R. (2014). Partisan power, economic coordination and variations in vocational training systems in Europe. European Journal of Industrial Relations, 20(1), 55-71. Ceron, A., & Negri, F. (2016). The “social side” of public policy: Monitoring online public opinion and its mobilization during the policy cycle. Policy & Internet, 8(2), 131-147. Joseph, N., Grover, P., Rao, P. K., & Ilavarasan, P. V. (2017). Deep analyzing public conversations: Insights from Twitter analytics for policy makers. In A. Kar et al. (Eds.) Digital nations – Smart cities, innovation, and sustainability. I3E 2017. Lecture Notes in Computer Science (pp. 239-250). Springer: Cham. Jussila, J., Boedeker, M, Jalonen, H., & Helander, N. (2017). Social media analytics empowering customer experience insight. A. Kavoura et al. (Eds.), Strategic Innovative Marketing. Springer Proceedings in Business and Economics (pp. 25-30). doi:10.1007/978-3-319-56288-9_4 Kwak, H., Lee, C., Park, H., & Moon, S. (2010). What is Twitter, a social network or a news media? Proceedings of the 19th international conference on world wide web (pp. 591-600). New York: ACM.