Study Introduction

Background

This research was driven by my past working experience and research interest in using PBL to substantiate deep learning. According to one project I have participated in, though teachers’ attitude toward PBL was mostly positive, some complained that PBL sometimes “looks messy and not productive.” This finding made me wonder how others, particularly teachers/educators/researchers, think about PBL. Twitter, in this case, provides a perfect platform to collect a larger quantities of data in a natural way (with little to nil social desirability bias), which cannot be accomplished by traditional research methodologies, such as surveys or interviews.

Research Purpose and Questions

This study intends to depict a brief image of people’s general opinion toward project based learning (PBL). This research is therefore driven by the following questions:

  1. What do people say about project-based learning (PBL) on Twitter?
  2. What key topics do people talk about PBL on Twitter?
  3. In general, what do Twitter users feel about PBL?
  4. What are the relationships between words (bigram) of the PBL tweets?

Methods

Data Source: With the keywords “PBL,” “project based learning,” and “project based inquiry” (case sensitive) with hashtag (#), I used the Twitter API to “listen to and analyze” (Twitter API) public conversations relevant to project based learning and try to grasp a picture out of it.

I selected 5 (status_id, user_id, text, screen_name, source, created_at) out of 91 variables to conduct the intended analysis. Due to the limited memory of my laptop, the number of tweets was reduced to 200 to run topic modeling analysis.

Data Processing: Tokenized, stemming, tidied text, document-term matrix.

Data Analysis: Word counts, tf-idf, document-topic probabilities, structural topic model (STM), Latent Dirichlet allocation (beta), and correlation (phi-coefficient).

Research Question Data Analysis
QR1. What do people say about project-based learning (PBL) on Twitter? unigram word count, tf-idf, quotes of top words based on word count and tf-idf
QR2. What key topics do people talk about PBL on Twitter? topic modeling (STM, LDA)
QR3. In general, what do Twitter users feel about PBL? unigram sentiment value analysis using AFINN, bigram analysis of negated words, quotes of top sentiment value words based on word count and sentiment value
QR4. What are the relationships between words (bigram) of the PBL tweets? correlation analysis using phi-coefficient

Targeted Audiences

Since PBL is a student-centered pedagogy which emphasizes on viewing students as learning agency and giving choice and voice to students through authentic projects, educational stakeholders such as teachers, school administrators, parents, educators, educational researchers, policy makers, and even students are natural audiences.

Considering this research only reports preliminary findings, interested readers are encouraged to further read research papers, books, or do research studies to dive deeper in this topic. They may otherwise find it interesting to compare and contrast what the authorities and the public say about PBL.

Findings

Words Used to Describe PBL by Word Count

Words Used to Describe PBL by tf-idf

What Did People Say About PBL on Twitter with the Top Words ‘check,’ ‘connecting,’ and ‘post?’

“Teachers! If you are keen to incorporate project-based learning into your #STEM lessons, but don’t know where to start, #CREST has some fantastic resources for you!Why not check out our secondary starter guide here.”

“Yes, let’s integrate news into the day on a regular base. Also, connecting to experts as often as possible.”

“Retweet this post and tell me why you want a copy of the book Project-Based Learning. I will select a winner and ship a copy out. Thanks.”

Top 5 Terms in Each STM Topic

Interactive Graph of Intertopic Distance Map and It’s 30 Most Salient Terms

Play with it Here: http://localhost:13267/session/file6c3c4d04a73/index.html

Top 5 Terms in Each LDA Topic

Users’ Overall Sentiment toward PBL

What Were Twitter Users Mostly Positive about on PBL?

“Do your students ever ask you: ‘When am I ever going to need this?’ If they do, consider rigorous #PBL units in middle and high school math as your answer to that age-old student question.”

“I’d like to listen, I truly believe we need project based learning, maker cultures and circular/civic economics at the earliest age, through fun and play. Not for ‘jobs’ market reasons, but to normalise fun/creativity/context/purpose as everyday not ‘special’ or siloed.”

“LoveLD - love this myth-busting post about Project Based Learning.”

What did users mean by ‘sucks,’ ‘fake,’ and ‘catastrophe?’

“Exactly! That budget would fund maybe 350 schools to engage local digital film companies to produce their own consent ads.. project based learning, content and stimluate local arts!Fkkrs. #milkshakevideo sucks.”

“Wildcatters combined two preexisting technologies to revolutionize the global energy. Delightfully this reduced coal use and dependence on corrupt regimes. Will we see a revolutionary combo to help K-12 recover from the COVID catastrophe?”

“I would argue that fake news is not any more prevalent than in the ‘olden days,’ it just distributes faster and I feel like people are less discerning in their critical analysis.”

Negated Bigrams

word1 word2 value n
no matter 1 3
not agree 1 2
not focused 2 1

Among a group of negative words “no, not, none, never, hardly, rarely,” only “no, not” were found in our dataset. Considering the total contribution of negated bigrams was 7, plus “no matter” is a rather neutral expression, the double negated effect and negated bigrams that need to be carefully examined in terms of sentiment analysis (Grimmer & Stewart, 2013) contributed an ignorable part to the overall sentiment value.

Words that Were Most Correlated with ‘environmental,’ ‘exploring,’ ‘math,’ ‘question,’ ‘resources,’ and ‘science’

Pairs of Words that Show at Least a 0.6 Correlation of Appearing

Conclusion and Discussions

Conclusion

  • Q1: What do people say about project-based learning (PBL) on Twitter?

    • People are using Twitter to share resources and call on participation related to PBL.

  • Q2: What key topics do people talk about PBL on Twitter?

    • Topic 1: What PBL is about and can bring about Project-based learning is a pedagogy endeavors to cultivating deep learning through projects. Terms such as “question,” “opportunities,” “share,” and “inquiry” in Topics 1, 2, 9 in LDA modeling and “curriculum” in Topics 3, 4, 6 in stm modeling are in correspondence with core elements of PBL.

    • Topic 2: In what domains PBL is deployed Topics 5, 7, 8 in LDA modeling and topics 1, 9 in stm lay out the domains covering linguistic, science, math, etc. However, STEM education (science, technology, engineering, and mathematics) seem to be predominant compared to other subject matters. topics 1, 9 in stm/gamma

    • Topic 3: Who is involved in PBL It is suggested that a wide range of parties are involved in the talk of PBL. Particularly, topic 10 in LDA modeling and Topics 2 & 13 in stm modeling showed that schools, tech organizations, and educational consulting corporations are all involved. It is worth noticing that PBL conversations on Twitter are mostly in the arena of k-12 education.

  • Q3: In general, what do Twitter users feel about PBL?

    • Generally speaking, most people on Twitter are feeling positive about PBL, believing PBL can bring ‘rigor’ and ‘fun’ in learning.

  • Q4: What are the relationships between words (bigram) of PBL tweets?

    • ‘adminchat,’ ‘cpchat,’‘mathchat,’‘suptchat,’ ‘elemchat’ seem to be active on Twitter to involve most of the PBL discussion about professional development and support.

    • PBL seem to be engaged with the topic of “inequity,” for instance, students of AP classes who are from low-income families.

Discussion

  • Limitations

    • Due to the limited memory of the author’s laptop, the data frame of this study is rather small (n=200), and therefore might not catch some features of people’s opinion about PBL on Twitter.

    • The findings only depict a brief picture of the phenomenon, further studies are needed to grasp deeper information out of it.

  • Next Steps for Analysis

    • Further studies using survey and interview engaging different participant subjects (e.g., teachers, students, coaches, administrators, principals, parents, etc) are particularly needed to expand our understanding of public opinions toward PBL (Cela, Sicilia, &Sánchez, 2015; He, 2012).

    • It is also interesting to do research comparing and contrasting the public (on social media) and academia’s views on PBL.

The Need for a High Quality PBL Framework

References

Cela, K.L., Sicilia, M.Á. & Sánchez, S. (2015). Social Network Analysis in E-Learning Environments: A Preliminary Systematic Review. Educ Psychol Rev 27, 219–246. https://doi.org/10.1007/s10648-014-9276-0

Grimmer, J., & Stewart, B. (2013). Text as Data: The Promise and Pitfalls of Automatic Content Analysis Methods for Political Texts. Political Analysis, 21(3), 267-297. doi:10.1093/pan/mps028

He, W. (2012). Examining students’ online interaction in a live video streaming environment using data mining and text mining. Computers in Human Behavior, 29(1), 90-102. https://doi.org/10.1016/j.chb.2012.07.020