Identifying Sources of Error - ACT Science

SECTION 1

Historical Context & Motivation

Science has always been a pursuit of accuracy, but throughout history, even brilliant researchers have been tripped up by experimental error—unintended factors that cause results to deviate from the true value. Understanding where errors come from is not just an academic exercise; it is the backbone of scientific credibility. On the ACT Science section, you will encounter passages that describe experiments, and the test frequently asks you to identify what went wrong, what could be improved, or what limits the conclusions a researcher can draw.

The story of error identification stretches back centuries. Early astronomers noticed that repeated observations of the same star yielded slightly different positions each time. Chemists discovered that impure reagents could ruin a reaction. Biologists realized that organisms respond differently under stress, making controlled conditions essential. Each of these revelations advanced our understanding of how to design reliable experiments—and how to critically evaluate them.

1610 • Galileo's Telescope Observations

Galileo recognized that lens imperfections in early telescopes distorted his observations of Jupiter's moons, prompting him to account for instrument limitations—one of the first documented discussions of systematic error.

1800s • Rise of Statistical Methods

Scientists like Carl Friedrich Gauss developed the concept of the normal distribution to describe how random errors scatter around a true value, giving researchers tools to quantify uncertainty.

1948 • Controlled Clinical Trials

The first randomized controlled trial for streptomycin treated tuberculosis patients under strict protocols, establishing the gold standard for minimizing bias and confounding variables in medical research.

2000s • Reproducibility Crisis

Researchers discovered that many published studies could not be replicated, sparking a global push to improve experimental design and transparently report all potential sources of error.

The central question this lesson addresses is straightforward but powerful: When you read about an experiment, how do you figure out where the results might have gone wrong—and what kind of error is responsible? Mastering this skill will help you on multiple ACT Science question types, from conflicting viewpoints to data representation passages.

SECTION 2

Core Principles & Definitions

Before you can spot errors on the ACT, you need a solid framework for the different categories of error that show up in scientific investigations. Every experiment aims to measure something as accurately and precisely as possible, but real-world conditions always introduce deviations. These deviations fall into distinct categories, each with its own causes and consequences.

Systematic Error

A consistent, repeatable flaw that shifts all measurements in the same direction. Think of a scale that always reads 2 grams too heavy—every measurement is off by the same amount. Systematic errors affect accuracy (closeness to the true value).

Random Error

Unpredictable fluctuations that cause measurements to scatter around the true value. These come from uncontrollable factors like slight vibrations, minor temperature changes, or natural variation among test subjects. Random errors affect precision (consistency of repeated measurements).

Human Error

Mistakes made by the experimenter, such as misreading a graduated cylinder, recording the wrong number, or failing to follow the procedure. On the ACT, human error is sometimes distinguished from systematic and random error because it can be eliminated through better technique or training.

Confounding Variables

Unmeasured factors that change alongside the independent variable, making it impossible to tell which variable actually caused the observed effect. A confounding variable threatens the validity of any cause-and-effect conclusion.

Sample Size & Selection Bias

A sample that is too small or not representative of the population introduces error into conclusions. If you test a drug on only five people, the results may not generalize. If you only test on adults, you cannot draw conclusions about children.

✦

KEY TAKEAWAY

Think of it like archery. Systematic error is like a misaligned sight—all your arrows land in a tight cluster, but the cluster is away from the bullseye. Random error is like a shaky hand—your arrows scatter unpredictably around the bullseye. The best experiment is one where the sight is aligned (no systematic error) and the hand is steady (minimal random error), so every arrow hits near the center.

SECTION 3

Visual Explanation — Accuracy vs. Precision

The relationship between accuracy and precision is one of the most frequently tested ideas related to error on the ACT. The diagram below uses the classic target analogy to show how systematic and random errors manifest differently in experimental data. Study each target carefully and notice how the pattern of dots changes.

Three targets illustrate the difference between accuracy and precision. The cyan dots are clustered near the center (ideal). The amber dots scatter around the center (random error). The pink dots cluster tightly but away from center (systematic error).

On the ACT, when a passage describes measurements that are consistently too high or too low, that signals systematic error—look for a calibration issue, a flawed instrument, or an uncontrolled variable that always pushes values in one direction. When a passage describes measurements that vary widely from trial to trial but average out near the expected value, you are seeing random error—the solution is to increase the number of trials so the average becomes more reliable.

SECTION 4

How Error Affects Experimental Results

While the ACT Science section does not require you to calculate error margins, it helps to understand the basic math behind error so you can interpret data tables and graphs more effectively. Two simple formulas capture the essence of accuracy and precision.

PERCENT ERROR (ACCURACY)

Percent Error = |Experimental Value − Accepted Value| ÷ Accepted Value × 100%

The absolute value bars (| |) mean you ignore whether the result was too high or too low. A low percent error indicates high accuracy. On the ACT, if a student's experimental result is far from the accepted value, the question may ask you to identify what systematic issue could explain the difference.

RANGE (SIMPLE PRECISION MEASURE)

Range = Maximum Value − Minimum Value

A small range among repeated trials indicates high precision. If the ACT shows a data table where one group's measurements have a much wider range than another's, a question might ask what caused the inconsistency—often pointing to random error or an uncontrolled variable.

Beyond these formulas, think about the mechanism of error in any experiment. Ask yourself three diagnostic questions when reading an ACT passage: (1) Could the measuring instrument itself be flawed? (2) Are the experimental conditions truly controlled, or could an outside factor be influencing the results? (3) Is the sample large enough and representative enough to support the conclusion being drawn? These three questions map directly onto the most common ACT error-identification prompts.

💡

ACT TEST TIP

When the ACT asks "Which of the following could be a source of error in the experiment?", eliminate answer choices that describe outcomes (e.g., "The temperature was too high") and look for choices that describe causes (e.g., "The thermometer was not calibrated before use"). The ACT wants you to identify the root cause, not the symptom.

SECTION 5

Classifying Error Types in Detail

On the ACT, you need to be able to quickly classify the type of error described in a passage and determine its impact. The diagram below provides a decision-tree flowchart that mirrors the thinking process you should use when you encounter an error-identification question.

Follow this decision tree when you encounter an error-identification question on the ACT. Start at the top: ask whether the error is consistent (systematic) or scattered (random), then consider the specific causes and fixes listed below each category.

Error Type	Common ACT Trigger Words	What to Look For
Systematic	"consistently higher," "all readings shifted," "always measured above"	Instrument calibration, environmental factor pushing all data one direction, procedural flaw
Random	"varied widely," "inconsistent results," "scatter in the data"	Uncontrolled fluctuations, small sample size, too few trials
Human	"misread," "recording mistake," "did not follow procedure"	Sloppy technique, parallax when reading a meniscus, math miscalculation
Confounding Variable	"could also explain," "another factor," "not controlled for"	An unmeasured variable changes alongside the independent variable, muddying cause-and-effect
Sample / Selection Bias	"only tested," "small group," "not representative"	Sample too small to generalize, sample not diverse enough, voluntary response

Quick-reference table for identifying error types using ACT-style language cues

SECTION 6

Worked Example — ACT-Style Passage Analysis

Let's walk through a realistic ACT-style scenario step by step. Read the mini-passage below, then follow the reasoning to identify the source of error.

ℹ

MINI-PASSAGE

A student investigated how temperature affects the rate at which sugar dissolves in water. She placed 10 g of sugar in 100 mL of water at five different temperatures (20 °C, 30 °C, 40 °C, 50 °C, and 60 °C) and timed how long the sugar took to dissolve completely. She performed three trials at each temperature. Her results showed that dissolution time decreased as temperature increased, as expected. However, her values at 40 °C were significantly higher than her classmates' values for the same temperature, while her other temperatures matched closely.

Question: Which of the following is the most likely source of error in the student's 40 °C trials?

A. The thermometer was broken and always read 5 °C too high.
B. The student used a different brand of sugar for the 40 °C trials.
C. Random fluctuations in room temperature affected all trials.
D. The student accidentally added only 80 mL of water in the 40 °C trials.

Step-by-Step Solution

1
Step 1 — Identify the Anomaly

The problem tells us that only the 40 °C data is off—it is significantly higher than classmates' values. The other temperatures match. This means the error is specific to one condition, not a global problem with the whole experiment.
2
Step 2 — Eliminate Global Errors

Choice A describes a broken thermometer that always reads 5 °C too high. If this were the case, all temperatures would be shifted, not just 40 °C. This is a systematic error, but it does not match the pattern described. Eliminate A.
3
Step 3 — Eliminate Random Error

Choice C suggests random room temperature fluctuations. Random error would cause scatter across all trials, not a consistent offset at just one temperature. Since the student ran three trials and all three at 40 °C were high, this does not fit a random pattern. Eliminate C.
4
Step 4 — Evaluate Remaining Choices

Choice B (different sugar brand) and Choice D (only 80 mL of water) both could explain an anomaly at only 40 °C. However, using a different brand of sugar would likely change the chemical composition, which is a confounding variable rather than a procedural error that would clearly increase dissolution time. Using less water (D) would mean less solvent, making it take longer for the sugar to dissolve—directly explaining why the dissolution times were higher.
5
Step 5 — Select the Best Answer

Choice D is the strongest answer because it identifies a specific procedural mistake (using the wrong volume of water) that directly and predictably causes the observed result (longer dissolution time at 40 °C only).

Answer: D — The student accidentally added only 80 mL of water in the 40 °C trials.

SECTION 7

Strengths & Limitations of Error Identification Strategies

Not every strategy for finding errors works equally well in every situation. Understanding the strengths and limitations of different approaches helps you choose the right mental tool when you encounter unfamiliar experiments on the ACT.

Strategy	Strengths	Limitations
Repeating Trials	Reduces impact of random error; averaging provides a more reliable result	Does not fix systematic error—if every trial has the same bias, more trials just reproduce the same bias more precisely
Using a Control Group	Helps isolate the effect of the independent variable; reveals confounding variables	Cannot eliminate random error; only useful if the control is truly identical except for the variable being tested
Calibrating Instruments	Directly eliminates one major source of systematic error	Only addresses instrument-based systematic errors; does not help with procedural or environmental biases
Increasing Sample Size	Makes results more generalizable; reduces selection bias; random fluctuations average out	More expensive and time-consuming; does not fix systematic errors or confounding variables
Blinding / Double-Blinding	Eliminates observer bias and placebo effects—critical in studies involving human subjects	Not applicable in many physics or chemistry experiments; adds complexity to study design

Comparison of common strategies for reducing experimental error

✦

KEY TAKEAWAY

Think of error reduction strategies like tools in a toolbox. A hammer is great for nails but useless for screws. Similarly, repeating trials is the right tool for random error, while calibrating instruments is the right tool for systematic error. On the ACT, match the fix to the type of error described—don't just pick the one that sounds most scientific.

SECTION 8

Connecting to Advanced Experimental Design

The error identification skills you build for the ACT are the same skills scientists use in professional research. In college-level science courses, you will encounter more formal treatments of error, including statistical measures like standard deviation, confidence intervals, and p-values. The table below compares the ACT-level understanding with the more advanced version, so you can see how this foundation connects to what comes next.

Concept on the ACT	Advanced Version in College
"Results varied widely" → random error	Calculate standard deviation (σ) to quantify how spread out the data is
"All values were too high" → systematic error	Perform calibration curves; apply correction factors to raw data
"Sample was too small" → unreliable conclusions	Power analysis determines the minimum sample size needed for statistically significant results
"Another factor could explain the result" → confounding variable	Multivariate regression analysis isolates the effect of each variable while controlling for others
Percent error calculation	Error propagation formulas track how uncertainty in each measurement compounds through calculations

ACT-level error concepts and their college-level counterparts

For now, you do not need to master any of these advanced techniques—the ACT will never ask you to calculate a standard deviation. However, understanding that your current knowledge is the foundation for these tools can help you appreciate why the ACT tests error identification so frequently. It is a skill that stays relevant throughout your scientific education.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A student measures the boiling point of water five times using the same thermometer and gets 103 °C, 103 °C, 102.5 °C, 103 °C, and 103.5 °C. The accepted boiling point of water at sea level is 100 °C. Which type of error best explains these results?

Show answer

The results are precise (clustered closely together) but not accurate (they are all about 3 °C too high). This pattern—consistent deviation in the same direction—is the hallmark of systematic error. The most likely cause is that the thermometer is improperly calibrated and reads approximately 3 °C above the true temperature.

Question 2: BASIC CALCULATION

A student experimentally determines the density of aluminum to be 2.85 g/cm³. The accepted value is 2.70 g/cm³. Calculate the percent error.

Show answer

Using the percent error formula: |2.85 − 2.70| ÷ 2.70 × 100% = 0.15 ÷ 2.70 × 100% ≈ 5.56%. This percent error indicates that the experimental value is about 5.6% higher than the accepted value. Since the experimental value is consistently above the true value, this suggests a systematic error—perhaps the student measured the volume too small (which inflates the density) or the mass too high (e.g., the sample was wet, adding extra mass).

Question 3: INTERMEDIATE

In an ecology experiment, a researcher counts the number of dandelions in ten randomly selected 1 m² plots in a field. The counts are: 12, 8, 15, 3, 22, 9, 14, 6, 18, and 11. The researcher concludes that the field has an average dandelion density of 11.8 per m². A critic argues this conclusion may be flawed. Identify two distinct sources of error in this study and explain how each affects the conclusion.

Show answer

First, random error is evident in the wide range of counts (3 to 22), reflecting natural spatial variation in where dandelions grow. This high variability makes the average of 11.8 less reliable; more plots would reduce the impact of this scatter. Second, selection bias could be a problem if the "randomly selected" plots were actually concentrated in one part of the field (perhaps near the edge where conditions differ). Even if the plots were truly random, ten plots from a large field may not be enough to capture the full range of conditions, leading to a sample that is not representative of the entire field.

Question 4: APPLIED

A pharmaceutical company tests a new headache medication. Group A (50 patients) receives the medication, and Group B (50 patients) receives a sugar pill placebo. Neither the patients nor the nurses administering the pills know which group receives which treatment. After two weeks, 70% of Group A reports fewer headaches, compared to 45% of Group B. However, it is later discovered that Group A patients were, on average, 10 years younger than Group B patients. Explain how this discovery affects the validity of the study's conclusions, and identify the specific type of error involved.

Show answer

The age difference between the two groups constitutes a confounding variable. Since younger patients may naturally experience fewer or less severe headaches, the observed improvement in Group A might be partly or entirely due to their younger age rather than the medication. The study used blinding correctly (neither patients nor nurses knew who got what), which is good for eliminating observer bias. However, the failure to match the groups by age means we cannot tell whether the medication or the age difference caused the result. To fix this, the company should randomly assign patients to groups in a way that balances age (stratified randomization), or they should statistically control for age in the analysis.

Question 5: CRITICAL THINKING

Two student lab groups measure the acceleration due to gravity using a pendulum. Group 1 reports values of 9.80, 9.79, 9.81, 9.80, and 9.80 m/s². Group 2 reports values of 10.5, 8.9, 9.2, 10.1, and 9.6 m/s². The accepted value is 9.81 m/s². Which group has the more reliable data, and which has the more accurate data? Propose one specific procedural change for each group that would most improve their results.

Show answer

Group 1 is both highly precise (range of only 0.02 m/s²) and highly accurate (average ≈ 9.80, very close to 9.81). Group 2 has a wider range (8.9 to 10.5), indicating low precision and significant random error. Their average is 9.66 m/s², which is further from the accepted value, so they are less accurate too. Group 1's data is more reliable because it is both precise and accurate. For Group 2, the most impactful improvement would be to standardize their measurement technique—for example, using a photogate timer instead of hand-timing the pendulum swings, which would reduce the random error caused by inconsistent human reaction times. For Group 1, their data is already excellent, but they could verify the length of the pendulum with a more precise ruler to ensure that their small systematic offset (9.80 vs. 9.81) is not due to a slight length measurement error.

SUMMARY

Lesson Summary

Summary

Every experiment is vulnerable to error, and the ACT Science section tests your ability to identify where things can go wrong. Systematic errors push all measurements in one direction and reduce accuracy; they are fixed by calibrating instruments and redesigning procedures. Random errors cause scatter and reduce precision; they are minimized by repeating trials and averaging results. Confounding variables threaten the validity of cause-and-effect claims and are controlled by using control groups and carefully isolating variables.

When answering ACT questions about error, use a systematic approach: first determine whether the error pattern is consistent (systematic) or scattered (random), then look for the root cause rather than the symptom. Check for human error like misreading instruments, sample size and selection bias that limit generalizability, and use the percent error formula to quantify how far results deviate from accepted values. Master these skills, and you will be well-prepared for any error-identification question the ACT throws at you.