A Systematic Review of Empirical Research on Graphing Numerical Data in K-12 STEM Education

This systematic review synthesizes 50 empirical studies on K-12 STEM education to demonstrate that while graphing instruction effectively enhances both graph construction and interpretation skills, students still encounter significant difficulties in creating, interpreting, and utilizing graphical representations of numerical data.

The Gist

Imagine you are trying to explain a complex story to a friend. You could write them a long, confusing letter, or you could draw a simple map. Graphing is that map. It's the skill of taking a pile of raw numbers (like temperatures, test scores, or stock prices) and turning them into a picture (a line, a bar, or a dot plot) that tells a clear story.

This paper is a massive "report card" for how schools teach kids (from kindergarten to 12th grade) to draw these maps. The authors didn't just look at one classroom; they dug through thousands of research studies to find the 50 best ones, creating a "systematic review" to see what actually works.

Here is the breakdown of their findings, explained with some everyday analogies:

1. The Big Picture: Why Do We Care?

Think of data as a jumbled box of LEGO bricks.

• Reading a graph is like looking at a finished LEGO castle. You can see the shape and the colors.
• Making a graph is like taking those loose bricks and actually building the castle yourself.

The researchers found that while most adults don't need to build LEGO castles in their daily jobs, learning to build them in school is crucial. It trains your brain to understand the world. If you can build the graph, you understand the data much better than if you just looked at someone else's finished castle.

2. How Do Teachers Teach This? (The "How-To")

The review looked at how different teachers tried to teach this skill. They found a few main ways:

• The "Do It Yourself" Method: Students get a list of numbers and have to draw the graph by hand with a pencil and ruler.
• The "Tech-Savvy" Method: Students use computers or tablets to click and drag the data into a graph.
• The "Hybrid" Method: A mix of both.

The Verdict: It doesn't matter too much which tool you use (pencil vs. computer). What matters is that the student is actively building the picture. The act of building it forces their brain to pay attention to the details, which helps them learn.

3. Does It Work? (The "Did It Help?" Test)

The researchers asked: Does teaching kids to draw graphs actually make them smarter at understanding data?

• Yes! It's like going to the gym. If you just watch someone lift weights, you don't get strong. But if you lift the weights yourself, your muscles grow.
• Students who practiced making graphs got better at reading graphs, too.
• It also helped them get better at other science skills, like forming hypotheses and arguing their points with evidence.

4. Where Do Students Get Stuck? (The "Speed Bumps")

Even though it works, it's not easy. The study found that students hit three main "speed bumps" when trying to build their LEGO castle:

• Bump #1: The Blueprint Errors (Construction Difficulties)

  • The Metaphor: Imagine trying to build a house but you put the door on the roof or the windows are upside down.
  • The Reality: Students often mess up the rules of the graph. They might put the wrong numbers on the wrong axes (like putting "Time" on the vertical line instead of the horizontal one) or make the scale look weird (like making one inch equal to 100 degrees instead of 10).

• Bump #2: The Translation Problem (Data Translation)

  • The Metaphor: You have a recipe written in English, but you need to cook using a French cookbook. You know the ingredients, but you can't translate the instructions.
  • The Reality: Students struggle to turn a table of numbers into a picture. They can read the numbers, but they don't know how to "translate" them into a dot on a line.

• Bump #3: The "It's Just a Picture" Trap (Theoretical Difficulties)

  • The Metaphor: Looking at a map of a forest and thinking, "Oh, that's just a drawing of trees," instead of realizing, "That drawing tells me where the bears live."
  • The Reality: Students often treat the graph as just a pretty picture rather than a tool for thinking. They might pick the wrong type of graph (like using a pie chart for something that needs a line graph) because they don't understand the story the data is telling.

5. The Takeaway for Teachers and Parents

The authors conclude that we need to stop just showing kids graphs and start making them build them.

• Don't just show the map; let them draw it.
• It's okay to make mistakes. The struggle of getting the scale wrong or picking the wrong graph type is actually where the learning happens.
• Connect the dots. Teachers need to help students realize that the graph isn't just a drawing; it's a tool to solve real-world mysteries, like predicting the weather or understanding why a plant grew taller.

In short: Graphing is like learning a new language. You can't just memorize the dictionary (reading graphs); you have to write sentences (making graphs) to truly speak it fluently. This paper confirms that while it's a tough skill to learn, the effort pays off by making students better thinkers in math, science, and everyday life.

Technical Summary

1. Problem Statement

Graphing—the construction of convention-based graphical representations of numerical data—is a fundamental competency in K-12 STEM education and a critical skill for data literacy in everyday life. While the ability to interpret graphs is widely studied, the specific process of constructing graphs from numerical data is less synthesized.

• The Gap: Despite graphing being a standard student activity, empirical research on it is fragmented, utilizing diverse study designs, analysis methods, and theoretical frameworks. No prior systematic review has synthesized findings regarding the implementation, effectiveness, and specific difficulties associated with student graphing in K-12 STEM contexts.
• The Need: Educators require a comprehensive overview to understand how graphing instruction impacts learning outcomes and what specific barriers students face, to better design curricula and interventions.

2. Methodology

The authors conducted a systematic literature review following the PRISMA 2020 guidelines.

• Search Strategy:
  • Databases: Scopus, ERIC, and PsycInfo.
  • Search Terms: Combinations of education terms (e.g., "educat*", "student*"), graphing terms (e.g., "graphing", "plotting"), and data terms (e.g., "data*", "variable*").
  • Timeframe: Initial search in March 2022; updated in April 2024.
  • Supplementary Search: Proceedings from seven major international conferences (e.g., AERA, ICOTS, ESERA) from 2019–2024.
• Inclusion Criteria:
  • Population: K-12 students (excluding university students).
  • Topic: STEM disciplines (Science, Technology, Engineering, Math).
  • Activity: Self-generated or partially completed graphs based on numerical data (excluding pure function graphs based on equations or provided graphs).
  • Type: Peer-reviewed empirical research (qualitative, quantitative, or mixed methods).
• Selection Process:
  • Initial search yielded 12,945 records.
  • After deduplication, title/abstract screening (using ASReview), and full-text review (using Rayyan), 50 studies were included.
• Data Extraction & Coding:
  • Two independent coders extracted data using a structured scheme covering: demographics, STEM topic, graph type, data collection method (self-generated vs. provided), guidance level, study design, theoretical framework, and reported difficulties.
  • Discrepancies were resolved through discussion to reach 100% agreement.

3. Key Contributions

This review provides the first synthesized evidence base on student graphing in K-12 STEM, offering:

1. A Taxonomy of Implementation: A detailed breakdown of how graphing is currently taught (manual vs. tool-based, guidance levels, data types).
2. Effectiveness Synthesis: An evaluation of graphing as an instructional method, distinguishing between skill acquisition and conceptual understanding.
3. Difficulty Classification: A structured categorization of student errors into convention-based difficulties (construction mechanics) and theoretical difficulties (conceptual understanding and interpretation).
4. Theoretical Contextualization: An analysis of the theoretical frameworks used (or omitted) in existing research, highlighting a lack of cognitive theory in many studies.

4. Key Results

A. Implementation Characteristics

• Topics: Mathematics was the most common context (n=22), followed by Physics, Biology, and Chemistry.
• Graph Types: Line graphs were the most frequently constructed (n=27), likely due to the prevalence of bivariate data.
• Methodology:
  • Manual vs. Tool: Manual graphing (n=29) was more common than tool-based (n=9), though 10 studies used a combination.
  • Data Source: Students often generated their own data (n=23) via experiments, while others were provided data (n=18).
  • Guidance: Most studies provided minimal or no specific graphing instruction (n=30), while 10 provided explicit instruction.
• Theoretical Basis: 23 studies lacked a specific theoretical background. Those that did cite frameworks included Generative Learning, ICAP (Interactive, Constructive, Active, Passive), and Ainsworth's DeFT framework.

B. Effectiveness of Graphing

• Positive Outcomes: Graphing instruction generally improved both graph construction and graph interpretation skills.
• Instructional Impact: Explicit instruction was found to be superior to unguided practice. Studies showed that graphing activities improved scientific process skills (e.g., hypothesis generation, argumentation) beyond just graphing mechanics.
• Manual vs. Tool: Two studies suggested manual graphing might be more effective than computer-generated graphs for deep conceptual understanding, though the evidence is limited by the age of some comparative studies.
• Developmental Trends: Graphing skills generally improved with age and grade level, correlating with mathematical understanding and cognitive development.

C. Student Difficulties

Out of 50 studies, 42 reported specific difficulties, categorized as:

1. Construction Difficulties (Most Frequent):
   • Scaling: Incorrectly scaling axes.
   • Variable Ordering: Plotting points at incorrect coordinates (e.g., swapping x and y).
   • Labeling: Failure to label axes or units correctly.
2. Theoretical Difficulties:
   • Conceptual Connection: Inability to link the graph to the underlying STEM concept (e.g., "graph-as-picture" errors where the graph is seen as a literal picture of the phenomenon rather than a symbolic representation).
   • Interpretation: Difficulty making predictions or inferring trends from the data.
   • Graph Type Selection: Choosing inappropriate graph types for the data structure.

• Interconnection: The review found a strong correlation between construction errors and theoretical misunderstandings; students who struggled with concepts often failed to construct the graph correctly.

5. Significance and Implications

• Pedagogical Value: The review confirms that graphing is not merely a mechanical skill but a generative learning activity that fosters deep conceptual understanding.
• Instructional Recommendations:
  • Explicit Instruction: Teachers should provide explicit guidance on graphing conventions, not just the underlying science.
  • Dual Focus: Instruction must address both the mechanics of graphing (scaling, labeling) and the theory (connecting data to concepts).
  • Iterative Process: A suggested pedagogical approach involves having students graph their initial perceptions of a phenomenon, then revising the graph based on measured data to bridge the gap between intuition and scientific evidence.
• Future Research: The authors call for more hypothesis-driven, theory-grounded studies, longitudinal research on skill development, and investigations into moderating variables (e.g., gender, prior knowledge) to refine instructional strategies.

Conclusion: Graphing is a high-value educational activity in K-12 STEM that improves both technical and conceptual competencies. However, significant barriers exist regarding conventions and conceptual mapping, necessitating targeted, explicit instruction that integrates graphing with scientific inquiry.