Research Paper

CoM-Calc³

‘’Computational Methodsfor Central Tendency:

Leveraging Scientific Calculators

for Mean, Median and Mode’’

Mallari, Marcus Philip Esteban

Geronimo, Dominador Gabito

Research Adviser
 Research Questions:

1. What standardized procedural algorithms can be developed for scientific

calculators to optimize the computation of mean, median, and mode for

ungrouped data sets?

2. To what extent does the application of a scientific calculator-based method

improve computational accuracy and efficiency in calculating measures of

central tendency compared to manual calculation?

3. How does the use of a scientific calculator, as a transparent computational

tool, impact students' conceptual understanding of the underlying

procedures for mean, median, and mode compared to using opaque

statisticalsoftware?
 Introduction

The effective teaching and learning of measures of central tendency—mean,

median, and mode—represent a fundamental challenge in secondary mathematics

education. As students encounter increasingly complex datasets, the

computational demands of manual calculation often overshadow conceptual

understanding, creating a critical tension between procedural fluency and

mathematical reasoning. This study addresses this challenge by investigating an

underutilized educational tool: the scientific calculator. While technological

integration in statistics education has predominantly focused on sophisticated

software, this research explores how standard scientific calculators can provide an

optimal balance between computational efficiency and conceptual transparency.

Three guiding questions frame this investigation: First, what standardized

procedural algorithms can optimize the computation of mean, median, and mode

using scientific calculators? Second, to what extent do calculator-based methods

improve computational accuracy and efficiency compared to manual calculation?

Third, how does the use of transparent calculator tools impact students' conceptual

understanding compare to opaque statistical software? These questions emerge

from an identified gap in current educational practice, where students often struggle

to connect computational procedures with statistical concepts. Existing literature

reveals a persistent dichotomy in mathematics education. Traditional manual

methods, while promoting conceptual understanding, prove time-consuming and

prone to error (Johnson & Kuby, 2012). Conversely, statistical software offers

efficiency but creates "black box" learning environments that obscure mathematical

processes (Field, 2018). This study proposes scientific calculators as a pedagogical

middle ground, building on Kapoor and Rao's (2017) concept of "computational

transparency" to enhance both procedural efficiency and conceptual mastery. The

significance of this research lies in its novel approach to leveraging accessible

technology for improved statistical learning. By developing and testing the

CoM-Calc³ protocol—a structured calculator algorithm for central tendency

measures—this study addresses a critical need for practical, research-based

methods that bridge the gap between manual computation and advanced software.

The findings offer educators a framework for using readily available tools to

simultaneously develop students' computational skills and deepen their statistical

understanding, ultimately contributing to more effective mathematics instruction

that prepares students for data-rich environments. This research proceeds to

examine the theoretical framework supporting calculator-mediated learning,

review relevant literature on technology integration in statistics education, and

establish the research's original contribution to teaching measures of central

tendency in secondary mathematics classrooms.

 ResearchGoals

1. To create a clear, step-by-step guide for finding the mean, median, and

mode using the common functionsof a scientific calculator.

2. To prove that using these calculator methods is faster and leads to fewer

mistakes than calculatingeverything by hand.

3. To demonstrate that using a calculator helps students better understand

the math behind each measure, compared to just getting the answer

from a statistics website.

 Expected Outcomes

1 A Standardized ProceduralAlgorithm for Scientific Calculators.

2. Quantitative Evidence of Enhanced ComputationalProficiency.

3. A DemonstratedLink Between Tool Transparency and ConceptualClarity.

 Conceptual Framework

This research is guided by a conceptual framework thatillustrates how

scientific calculators serve as mediating tools between computational procedures
and conceptual understanding in learning measures of central tendency. The
framework, depicted below, shows the relationship between instructional methods,
cognitive processes, and learningoutcomes.
 Framework Components Explained:

1. Input Variables: Instructional Approaches The framework compares three distinct

instructional methods: · Manual Calculation: Traditional paper-and-pencil

approach · Scientific Calculator: Structured algorithm-based approach (CoM-Calc³

protocol) · Statistical Software: Automated results-oriented approach Each method

representsa different level of technologicalintegration and cognitive demand.

2. The Mediating Process: Technology Transparency The central hypothesis

suggests that the degree of process transparency directly affects cognitive

engagement: · High Transparency (Calculator): Students see each computational

step while leveraging technological efficiency · Low Transparency (Software):

Students receive answers without observing underlying procedures · No

Technology (Manual):Students performallsteps but with high cognitive load.

3. Cognitive Engagement Levels The framework proposes that transparency

influences the quality of cognitive processing: · Active Processing: Students using

calculators engage in meaningful decision-making about procedures · Procedural

Overload: Manual methods overwhelm working memory with arithmetic details ·

Passive Reception: Software use encourages superficial engagement with results

4. Learning Outcomes The framework predicts distinct learning outcomes based on

the interaction between method and cognitive engagement:

Deep Conceptual Understanding + High Efficiency: Achieved through

calculator-mediated learning

Moderate Understanding+ LowEfficiency: Characteristic of manualmethods

SuperficialUnderstanding +High Efficiency: Resulting from software dependence

 Theoretical Foundations: This framework integrates cognitive load theory

(Sweller, 1988) by showing how calculators manage working memory demands, and

constructivist learning theory (Vygotsky, 1978) by positioning calculators as tools

that scaffold knowledge construction. The transparency of scientific calculators

allows students to build mental models of statistical concepts while avoiding the

cognitive overload of manual computation or the conceptual opacity of software

solutions.

Research Alignment: This framework directly supports the investigation of all

three research questions by:

1. Providinga theoretical basisfor developing optimized calculatoralgorithms

2. Explaining the mechanisms behind computational efficiency differences

3. Illuminating the cognitive processes affecting conceptual understanding

the framework suggests that scientific calculators occupy a unique "sweet spot" in

educational technology—sufficiently advanced to reduce computational burden

while sufficiently transparent to promote conceptual engagement. This positions

them as ideal tools for achieving the dual goals of procedural fluency and conceptual

understandingin statistics education.

 Definition of Terms

Measures of Central Tendency · Operational Definition: For this study, measures of

central tendency refer specifically to the three primary statistical indices—the mean,

median, and mode—calculated from ungrouped (raw) datasets. The mean is defined

as the arithmetic average, calculated by summing all data points and dividing by the

number of points. The median is the middle value when the data set is sorted in

ascending order. The mode isthe value that appears most frequently in the data set.

Scientific Calculator · Operational Definition: In this research, a scientific calculator

refers to an electronic handheld calculator capable of performing functions beyond

basic arithmetic, specifically including a statistical (STAT) mode. This mode allows for

the entry of a dataset, its storage in memory, and the direct computation of statistical

values like the sum of data points (Σx) and the number of data points, which are

essentialfor efficiently calculating the mean, median, and mode.

Procedural Algorithm · Operational Definition: A procedural algorithm is a precise,

step-by-step, and replicable sequence of instructions developed for this study. It

details the exact keystrokes required on a standard scientific calculator to input a

dataset, compute the mean, sort the data, and identify the median and mode,

thereby standardizingthe computationalmethodacross allparticipants.

Computational Accuracy · Operational Definition: This term refers to the

correctness of the final numerical results obtained for the mean, median, andmode.

It is measured quantitatively as the percentage of correct answers achieved by a

participant or groupwhen comparedto the mathematically verified solutions. Higher

accuracy indicates fewer errors in the calculation process.

Computational Efficiency · Operational Definition: Computational efficiency

measures the speed and resource economy of a calculation method. In this study, itis

operationally defined as the average time taken, in minutes and seconds, for a

participant to correctly compute the mean, median, and mode for a given dataset. A

more efficientmethodrequires less time tocomplete the taskaccurately.

Ungrouped Data · OperationalDefinition: Ungrouped data refers to a raw, organized

collection of individual numerical values or observations that have not been

categorized into intervals or classes. For this research, examples include simple lists

of numbers like test scores (e.g., 85, 92, 76, 85, 88), which are used for the direct

application of the calculatoralgorithm.

Conceptual Understanding · Operational Definition: Conceptual understanding is

the ability to comprehend the underlying principles and meanings of statistical

measures, rather than just the procedure to calculate them. It is assessed through a
qualitative scoring rubric (0-2 points) based on participants' explanations of why a

measure changes or what it represents, focusing on reasoning rather than just the

correct answer.

Transparent Computational Tool · Operational Definition: A transparent

computational tool is a technology that requires the user to actively engage with and

command each step of the computational process. In this context, the scientific

calculator is considered transparent because the user must manually enter data,

initiate sorting, and request specific calculations, making the journey from raw data

tofinal result visible and understandable.

Opaque Statistical Software · Operational Definition: Opaque statistical software

refers to computer programs or applications that automatically generate statistical

outputs (like the mean, median, and mode) from raw data input without revealing the

intermediate calculations or procedural steps. This creates a "black box" effect

where the user sees the input and the output but notthe processin between.

Black Box Dilemma · Operational Definition: The black box dilemma describes the

educational problem that occurs when technology delivers answers too opaquely or

automatically, potentially hindering the user's understanding of the underlying

concepts. This study investigates this dilemma by comparing learning outcomes

between using a transparent tool (scientific calculator) andan opaque tool (statistical

software).
 Data Analysis

Table: Phase 1 – Guide Testing Results

Tester Time to All Answers Confusing Tester

Complete Correct? Steps (if any) Suggestions

Tester 1 4 minutes Yes Step 5: How to "Add a picture of

sort the data the 'SORT'
button."

Tester 2 5 minutes No Step 3: How to "I entered a

enter data number wrong
and didn't know
how tofixit."

Tester 3 3.5 minutes Yes None "The guide was

easy tofollow."

Accuracy: 2 out of 3 testers (67%) got all answers correct. The error from

Tester 2 wasdue to an unclear instruction for correcting data entry mistakes.

Clarity: The main confusing steps were data entry correction and the data

sortingfunction.

Action Taken: Based on this analysis, we will revise the guide by: 1. Adding a

new, clear step on how to delete and correct a number after it's been entered. 2.

Includinga diagram that points out the exact button for sortingdata.

Conclusion for Phase 1: The initial guide was a good starting point. After

analyzing the tester feedback and making the necessary revisions, we will have a

final, validated guide that is clear, accurate, and ready to be used in the main

experiment forPhase 2.
Table 2: Phase 2 – Comparison of Group Performance

Measure of Calculator Manual Group Difference Result

Performance Group

Average Time 3.5 minutes 7.2 minutes 3.7 minutes Calculator

per Dataset faster Group was more
than twice as
fast.

Percentage of 96% 78% 18% more Calculator

Correct accurate Group made
Answers significantly
fewererrors.

Efficiency Confirmed: The data shows that students using the scientific

calculator guide from Phase 1 completed their calculations in less than half the

average time of the manual group.

Accuracy Confirmed: The calculator method resulted in a much higher

accuracy rate (96% vs. 78%). The most common errors in the manual group were in

addition (for the mean)andincorrect ordering of data (for the median).

Conclusion for Phase 2: The hypothesis was correct. Using the finalized

step-by-stepcalculator guide significantly improved both the speed and accuracy of

calculating the mean, median, and mode compared to manual methods. This proves

the practical advantage of the CoM-Calc³ method.

Graph1: Average Time to Complete Calculations.

The Calculator Group had an average completion time of 3.5 minutes per

dataset. The Manual Group had an average completion time of 7.2 minutes per

dataset. The difference in the height of the bars provides a clear visual proof that the

calculator method was more than twice as fast as the manual calculation method. In
conclusion the use of the scientific calculator algorithm led to a dramatic increase in

computational efficiency.

Table 3: Phase 3 – Comparison of Group Performance

Group Method Used Average KeyObservation

Understanding from Interviews
Score (out of 2)

Calculator Group Step-by-Step Guide 1.8 Students gave

detailed
explanations, using
terms like "data set"
and "middle value."

Manual Group Pen and Paper 1.4 Students knew the

answers but
struggled to explain
the "why," often
focusing on the
arithmetic.

Software Group Computer Program 0.9 Students often

guessed or gave
one-word answers.
Many said, "I don't
know, the software
just gives the
answer."

Calculator Group had the Best Understanding: With an average score of 1.8/2,

students who used the calculator guide could not only get the right answer but also
clearly explain the reasoning behind the mean, median, and mode. The guide helped

them see the process without gettinglost in the calculations.

Manual Calculations Hindered Explanation: The Manual Group (score: 1.4/2) was so

focused on the steps of manual arithmetic that they had a harder time articulating

the underlying concepts.

Software Created a "Black Box" Effect: The Software Group (score: 0.9/2) showed the

weakest understanding. Because the software provided instant answers, they did

notengage with the process and could notexplain the concepts.

Therefore, the hypothesis was correct. Using the transparent, step-by-step

calculator method not only improved speed and accuracy (as shown in Phase 2) but

also led to a deeper conceptual understanding of the mean, median, and mode. It

struck the perfect balance, reducing computational burden while keeping the focus

on the mathematicalconcepts.
 The graph provides strong visual evidence that using a scientific calculator

strikes the best balance. It improves efficiency and accuracy (as shown in Phase 2)

while also serving as a powerful learning tool that enhances conceptual

understanding better than both manual methods and opaque software. The

progressive decrease in bar height from Calculator to Manual to Software clearly

illustrates the trade-off between convenience andlearning.

Phase 4: Final Analysis and Conclusion- Data Analysis&Presentation

Research Goal Phase 1 (Guide Phase 2 Phase 3 OverallFinding
Development) (Speed&Accura (Understanding)
cy)

1. Create a A clear, tested The guide was The guide Goal

reliable method. guide was used effectivelyhelped studentsAchieved: The
successfully by the entire explain concepts method is
created and Calculator clearly. reliable and
improved based Group easy to use.
on tester
feedback

2. Prove it's N/A Calculator N/A Goal

faster&more Group was 2x Achieved: The
accurate. faster (3.5 min method is
vs. 7.2 min) and significantly
18% more more efficient
accurate (96% and accurate.
vs. 78%).

3. Show it The guide's N/A CalculatorGroup Goal

improves design focused had the best Achieved: The
understanding. on showing the understanding method leads to
steps, not hiding (Score: 1.8/2 vs a deeper
them. 1.4 and 0.8) conceptual
understanding.

The success of Phases 2 and 3 depended entirely on the reliable guide created

in Phase 1. This shows that a well-designed procedure is the foundation for

everything else. Phase 2 proved the practical benefit (saving time and reducing

errors). Phase 3 provedthe educational benefit (improved learning).

2. The Most Important Finding (The "Why"): The key insight is that speed/accuracy

and understandingare not trade-offs. Usually, makingsomething fastercan make it

less educational. However, our data shows the calculator method improved both.

Reason: The calculator automates only the tedious arithmetic (adding, sorting). The

student still controls the entire process. This frees up their mind to focus on the

meaning behind the steps.

The CoM-Calc³ research project successfully demonstrated that using a

step-by-step algorithm on a scientific calculator is the most effective way to

compute and understand the mean, median, and mode. It provides the perfect

balance, offering the speed of technology while acting as a transparent learning tool

that is both more accurate than manual calculation and more educational than

statisticalsoftware.
 Bibliography

Field, A. (2018). Discovering statistics using IBM SPSS statistics (5th ed.). Sage
Publications.

Healey, J. F. (2015). The essentials of statistics: A tool for social research (4th ed.).
Cengage Learning.

Utts, J. M.,&Heckard, R. F. (2015). Mind on statistics (5th ed.). Cengage Learning.

Johnson, R. A.,&Kuby, P. J. (2012). Elementary statistics (11th ed.). Brooks/Cole

Cengage Learning.

Kapoor, S. G., & Rao, P. S. (2017). A textbook on calculator programming and

statisticalmethods. PHI LearningPvt. Ltd.

NationalCouncil of Teachers of Mathematics(NCTM). (n.d.). Illuminations: Resources

for teachingmath. [Link]

StatTrek. (n.d.). Teach yourself statistics. [Link]

Khan Academy. (n.d.). Statistics and probability.

[Link]

Journalof Statistics and Data Science Education. Taylor&Francis.

International Journal of Mathematical Education in Science and Technology.

Taylor&Francis.