Unit 2 Independent Study: Analyzing Sentiment

Purpose

Educational standards like the Common Core State Standards (CCSS) and Next Generation Science Standards (NGSS) have been widely debated, with varying perceptions among educators, policymakers, and the public. This analysis is focused on twitter data pertaining to CCSS and NGSS from January to May 2020.

Research Question

How did sentiment toward CCSS and NGSS change from January to May 2020?

Methods

Data Selection & Collection

The dataset consists of tweets posted between January and May 2020 that contain keywords related to CCSS and NGSS, including:

• CCSS-related keywords: “#ccss”, “common core”

• NGSS-related keywords: “#ngsschat”, “ngss”

Data Preparation

Before analysis, several preprocessing steps were applied:

• Text cleaning: Removed URLs, mentions (@user), hashtags, special characters, and excessive whitespace.

• Tokenization: Split tweets into individual words.

• Stopword removal: Removed common words (e.g., “the,” “and,” “is”) that do not contribute to sentiment.

Sentiment Analysis Techniques

The NRC Emotion Lexicon was used to classify words into eight primary emotions and positive/negative sentiment categories:

• Emotions Tracked: Anger, Anticipation, Disgust, Fear, Joy, Sadness, Surprise, Trust

• Polarity: Positive vs. Negative

This approach allows for a deeper emotional analysis compared to traditional positive/negative classification methods like VADER.

Analytical Techniques

• Emotion Distribution Analysis: Compared which emotions were most associated with tweets about CCSS vs. NGSS.

• Sentiment Polarity Over Time: Tracked how positive and negative sentiment changed month by month.

Words tweeted frequently about either NGSS or CCSS

This word cloud, generated from tweets discussing CCSS and NGSS from January to May 2020, provides insight into the key topics of online discourse surrounding these education standards. The prominence of “math” as the most frequently mentioned term aligns with CCSS’s strong association with mathematics education, reflecting its central role in public discussions. Other frequently occurring terms, such as “science,” “students,” “education,” “school,” and “standards,” indicate broader conversations about curriculum and learning. Additionally, words like “teachers,” “teaching,” “learning,” and “curriculum” suggest that many tweets focus on instructional strategies and educators’ experiences with implementing these standards. Analyzing these trends helps contextualize shifts in sentiment toward CCSS and NGSS over the selected time period.

Findings

The following sentiment analysis uses the NRC sentiment lexicon.

The bar graph displays the total count of words associated with different sentiment categories (anger, anticipation, disgust, fear, joy, negative, positive, sadness, surprise, and trust) across all tweets related to CCSS and NGSS. The negative sentiment category has the highest word count, followed by positive sentiment, while other emotions like anger, trust, and fear also appear prominently. This indicates that discussions about these educational standards contain a significant amount of negative language.

Sentiment for CCSS Over Time

The line graph for CCSS tracks how frequently sentiment words appear over time from January to May 2021. Positive sentiment is consistently the highest, with negative sentiment remaining lower in comparison. Other emotions such as anger, trust, and sadness show relatively stable trends. While the bar graph suggested a high frequency of negative words overall, the line graph does not show large spikes in negativity over time.

Sentiment for NGSS Over Time

The NGSS sentiment trend graph also shows that positive sentiment dominates, but at a lower level compared to CCSS. The trust category is more prominent in NGSS discussions, while negative sentiment appears even less frequently than in CCSS tweets. This suggests that NGSS-related conversations may be less contentious and more neutral or supportive than CCSS discussions.

Summary

The bar graph highlights the overall presence of negative words, whereas the line graphs provide a more time-sensitive perspective, showing that positive sentiment consistently appears more frequently over time. The higher trust levels in NGSS tweets also suggest that it is viewed more favorably than CCSS, which appears to be a more polarizing topic in public discussions.

Discussion

Strengths of the Analysis

• Comprehensive Sentiment Analysis: Used the NRC lexicon to categorize tweets into multiple emotions, providing deeper insights beyond just positive/negative sentiment.

• Comparative Insights: Allowed for a direct comparison between CCSS and NGSS, showing differences in public perception.

• Time-Based Trends: The line graphs helped track how sentiment evolved over time, offering a more dynamic understanding of public discourse.

• Large Dataset from Social Media: Twitter provided a rich dataset of real-time opinions, making the analysis relevant for understanding public sentiment.

Weaknesses of the Analysis

• Word-Level Sentiment Analysis: NRC does not account for context, sarcasm, or negation, leading to potential misclassifications (e.g., “great” could be positive even in sarcastic tweets).

• Discrepancies Between Graphs:

  • Bar graph shows high negative sentiment due to total word counts.

  • Line graphs show dominant positive sentiment over time, suggesting tweets contain mixed emotions rather than being overwhelmingly negative.

• Limited timeline for the dataset

Uses of the Analysis

👩‍🏫 For Educators & Policymakers:

• CCSS shows high negative sentiment, indicating ongoing public concerns that may need better communication or policy adjustments.

• NGSS has higher trust and lower controversy, suggesting a more favorable public perception and potentially a better model for implementing education standards.

📊 For Researchers & Analysts:

• Findings highlight the importance of tracking sentiment trends to understand how educational policies are perceived over time.

• Can help identify key moments when public perception shifts

📰 For Public Engagement & Advocacy:

• Understanding negative sentiment in CCSS discussions can help education leaders address misconceptions, clarify policies, and improve public trust.

• Insights into NGSS trust levels can support advocacy for continued adoption and improvements in science education.

Future Improvements & Next Steps

• Expand the Timeframe: Analyze tweets over multiple years to capture long-term shifts in sentiment.

• Using different sentiment lexicons to analyze tweets.

Correlate Sentiment with Major Events:

• Compare sentiment trends with policy announcements, education reform debates, and legislative changes.

• Helps determine what influences public opinion shifts the most.

Conclusion

• CCSS discussions contain high negative sentiment, but positive sentiment dominates over time, possibly due to the NRC lexicon’s classification of certain words as positive.

• NGSS has more trust and less controversy, suggesting a more favorable public perception compared to CCSS.

• Future studies should refine sentiment classification, expand datasets, and incorporate multiple platforms to create a more accurate and comprehensive analysis.