Praxis 5572: Scientific Evidence and Engineering in Earth Science
Praxis 5572 integrates scientific and engineering practices with Earth science knowledge.
- Study how questions, models, measurements, error analysis, data, and design criteria support a conclusion.
- Separate observations from inferences, evaluate whether evidence fits a claim, and identify how tools and model limitations affect what can be known.
On this page6 sections
- Praxis 5572 begins with Nature and Impact of Science and Engineering, approximately 15 percent of the assessment. That category includes the nature of scientific knowledge, investigations, data, engineering design, technology, safety, resources, environmental effects, hazards, and applications of Earth and space science in daily life. These practices also cross into other categories. ETS says half or more of test questions integrate one or more Science and Engineering Practices.
- Accuracy is closeness to an accepted or true value; precision describes repeatability or spread. Random error varies in unpredictable ways and can reduce precision. Systematic error shifts values consistently and can reduce accuracy. Repeating measurements can help characterize random variation, but it will not fix a sensor with an incorrect calibration. Identify the error source before proposing a repair.
- Different data displays answer different questions. A topographic map shows elevation through contours; a geologic map represents surface units and structures; a cross section displays a vertical slice; a seismogram records motion over time; a weather map depicts conditions over a region. Check projection and scale before estimating distance. Make the data display part of the evidence chain.
- Engineering design starts with a problem, criteria, and constraints. Criteria describe what counts as success; constraints may include cost, time, materials, safety, maintenance, and environmental impacts. A solution is tested, compared against criteria, and improved. “Best” has no meaning until the criteria and constraints are explicit.
- Scientific decisions can affect communities and ecosystems. Evidence can estimate pollution, resource availability, climate patterns, or hazard probability, while a policy decision also weighs values and tradeoffs. Separate a measurement from a judgment about acceptable risk. A well-supported answer explains the evidence and uncertainty before stating a design or management choice.
- Worked example: a group measures streamflow before and after a storm and reports a larger peak downstream. The result supports a measured increase at that site and time. It does not establish that pavement alone caused the peak. Useful additional evidence includes rainfall, upstream conditions, soil saturation, drainage, repeated storms, and comparable catchments.
Praxis 5572 begins with Nature and Impact of Science and Engineering, approximately 15 percent of the assessment. That category includes the nature of scientific knowledge, investigations, data, engineering design, technology, safety, resources, environmental effects, hazards, and applications of Earth and space science in daily life. These practices also cross into other categories. ETS says half or more of test questions integrate one or more Science and Engineering Practices.
Observation records what a person or instrument directly detects; inference offers an explanation based on those observations and a model. The distinction matters in geoscience because many systems cannot be observed directly over long periods. A seismogram records wave arrivals; a model of the interior helps infer layer properties. A satellite measures energy in selected bands; researchers interpret the signal to infer a surface or atmospheric feature.
A scientific question should be answerable with evidence. A hypothesis is a proposed explanation or prediction that can be investigated. A test plan identifies relevant variables, how they will be measured, what comparison is useful, and which result would challenge the idea. Not every field study can control every condition, but clear design helps separate what the evidence supports from what remains uncertain.
Independent and dependent variables are defined by the question and design. A controlled factor is kept consistent to make a comparison interpretable. If a study compares stream temperatures above and below a discharge, researchers need comparable times, calibrated sensors, repeated measures, and information about other inflows. A single difference establishes a pattern at measured sites, but not necessarily its cause.
Accuracy is closeness to an accepted or true value; precision describes repeatability or spread. Random error varies in unpredictable ways and can reduce precision. Systematic error shifts values consistently and can reduce accuracy. Repeating measurements can help characterize random variation, but it will not fix a sensor with an incorrect calibration. Identify the error source before proposing a repair.
Units and scale are part of reasoning, not decoration. Carry units through calculations, convert with equivalent ratios, and check that the final dimension matches the requested quantity. Rates need compound units such as meters per year. Scientific notation and significant figures support readable magnitudes and precision. A map scale or graph interval can change the apparent size of a pattern.
Graphs and tables should be read from title to axes, units, legend, and scale. Decide which variables are shown and whether the question asks for a value, trend, comparison, relationship, or inference. A correlation may support a relationship without identifying a cause. Extrapolation beyond measured values is less secure than interpolation within them.
Different data displays answer different questions. A topographic map shows elevation through contours; a geologic map represents surface units and structures; a cross section displays a vertical slice; a seismogram records motion over time; a weather map depicts conditions over a region. Check projection and scale before estimating distance. Make the data display part of the evidence chain.
Models simplify systems to make them understandable or testable. A plate diagram may show relative motion but omit scale, velocity, and multiple forces. A water-cycle diagram may emphasize reservoirs but leave out residence times. A solar-system model may not preserve actual distances. Ask which features matter to the question and which are left out.
Model limitations do not make a model useless. A model can explain a pattern or generate a prediction when its assumptions fit the task. Models change when new evidence exposes a weakness or when a better measurement becomes available. Use a model to reason, then compare its predictions with observations and revise if they conflict.
Technologies extend measurement. Seismometers detect ground motion; satellites observe reflected or emitted radiation; radar and sonar map surfaces or subsurface features; GPS helps estimate location and motion; telescopes gather light from distant objects. Instruments have resolution, calibration, coverage, and processing limits. Technological output still requires interpretation and quality checks.
Engineering design starts with a problem, criteria, and constraints. Criteria describe what counts as success; constraints may include cost, time, materials, safety, maintenance, and environmental impacts. A solution is tested, compared against criteria, and improved. “Best” has no meaning until the criteria and constraints are explicit.
Earth science engineering includes hazard reduction, water management, resource use, pollution control, and infrastructure. A levee may reduce flood exposure in one area while changing flows elsewhere. A slope stabilization design has maintenance needs and ecological effects. Evaluate performance and tradeoffs for the specific location rather than assuming one technology solves every version of a problem.
Natural hazards can include earthquakes, volcanoes, landslides, floods, tsunamis, droughts, hurricanes, and wildfires. Hazard describes a potential physical event; risk also depends on exposure and vulnerability. Monitoring, warnings, land-use planning, building codes, and public communication can reduce consequences but cannot remove all risk. Distinguish prediction from probability and preparedness.
Scientific decisions can affect communities and ecosystems. Evidence can estimate pollution, resource availability, climate patterns, or hazard probability, while a policy decision also weighs values and tradeoffs. Separate a measurement from a judgment about acceptable risk. A well-supported answer explains the evidence and uncertainty before stating a design or management choice.
Safe work protects people and data. Field and lab plans should consider equipment condition, sample labels, weather, terrain, water, protective procedures, storage, and disposal. Contamination can compromise results as well as create a safety problem. The appropriate procedure depends on the specific material and task, so do not invent a universal precaution when the scenario provides details.
Question prompts may ask you to identify what additional evidence is needed. Choose data that would distinguish competing explanations. If two models predict different patterns, collect a measurement where those predictions diverge. More data are not automatically better if they repeat the same bias or fail to measure the relevant variable.
Worked example: a group measures streamflow before and after a storm and reports a larger peak downstream. The result supports a measured increase at that site and time. It does not establish that pavement alone caused the peak. Useful additional evidence includes rainfall, upstream conditions, soil saturation, drainage, repeated storms, and comparable catchments.
Worked example: a team selects between two water-treatment designs. One removes more sediment but uses more energy; the other costs less but may require frequent maintenance. Compare the designs against stated water-quality goals, energy use, budget, maintenance capacity, and waste effects. The scientifically “best” result depends on the engineering criteria.
When you miss a practice question, classify the cause as content, units, graph reading, investigation design, model limits, or inference. A targeted correction may be a unit drill, a model comparison, or a new data set. Repeating the same question without understanding the error can improve recall of the answer while leaving the underlying skill unchanged.
The companion organizes Science and Engineering Practices around asking questions, developing and using models, planning investigations, analyzing data, using mathematics, constructing explanations, engaging in argument from evidence, and obtaining or communicating information. These are not eight unrelated vocabulary terms. A prompt may combine several, such as using a model to generate a prediction and comparing it with measurements.
When comparing explanations, ask what evidence each one predicts. A stronger explanation accounts for more observations with fewer unsupported assumptions. If two explanations both fit current data, identify a measurement that could distinguish them. “More research is needed” is less useful than naming the variable, location, time, or instrument that would reduce uncertainty.
Sampling design affects what a data set represents. A few measurements from one accessible site may not represent a whole watershed, region, or season. Random sampling can reduce selection bias, while stratified sampling can ensure important environments are covered. Recognize how sample location and timing limit conclusions before interpreting a reported average.
Precision and significant figures should match instrument resolution. Reporting many decimal places does not make a measurement more accurate. If two readings differ by less than the uncertainty, the difference may not be meaningful. Use the uncertainty information in a prompt instead of treating every displayed digit as exact.
Engineering optimization means improving a solution against stated criteria, often under competing constraints. A design may reduce flood risk but require higher cost or alter habitat. Compare alternatives on a common basis and include maintenance or lifecycle effects where relevant. One impressive performance metric may not settle the full decision.
Life-cycle thinking follows a material or technology from extraction through use and disposal. Earth science decisions about minerals, energy, or water can have effects at each stage. A resource that is renewable on one timescale may be depleted locally if use exceeds replenishment. State the relevant spatial and temporal scale.
Environmental evidence may include baseline measurements, repeated monitoring, and comparison sites. A before-after comparison can be confounded by other changes; a control site can help, but it must be sufficiently comparable. Identify what changed besides the factor of interest and whether the data establish a causal effect.
Risk communication should express probability and consequence clearly. A hazard map may show zones of elevated likelihood or potential intensity, not a guarantee that an event will occur at a particular date. Engineers and agencies combine geologic evidence, monitoring, exposure, and vulnerability when planning mitigation.
For scientific models, distinguish input, process, output, and assumption. A climate model may use emissions and physical processes to estimate a range of future conditions; a tectonic model may relate boundary motion to stress and landforms. A mismatch between prediction and observation can reveal a missing process, poor parameter, or measurement issue.
Use a deliberate routine for multiple-choice data items: restate the question, note the relevant variables, identify the pattern, check units and scale, and compare each choice with the evidence. Distractors often describe a true concept that does not match the displayed data or claim more causation than the design permits.
For engineering tasks, explain why a choice meets a criterion rather than just naming a preferred technology. If a prompt prioritizes minimizing energy use, an option with the highest treatment capacity may not be best. If public safety is the primary constraint, cost alone may not decide. The wording defines how performance should be judged.
A source can contain reliable measurements and still be unsuitable for a particular question. Check whether the data are current enough, cover the relevant region, and measure the needed variable. A global average cannot answer every local question, and a short record may not resolve a long-term pattern.
Keep diagrams clear when you study. Label system boundaries, variables, direction of flow, and evidence. Use separate arrows for matter and energy when both move. A tidy representation helps you see whether a conclusion depends on a hidden assumption, such as treating an open system as closed.
Practice explaining the difference between a scientific explanation and an engineering decision. The explanation describes a process supported by evidence. The decision selects a solution according to criteria and constraints. A question may require both, but they are not interchangeable; a design recommendation needs more than a causal description.
Before moving on, test yourself with a new context. Interpret an unfamiliar earthquake map, compare two water-treatment designs, or evaluate a graph of stream temperature. Explain the evidence and one limitation without looking at notes. That transfer check is a better readiness signal than rereading the same worked example.
If the same measurement is collected by different instruments, compare calibration, resolution, placement, and procedure before treating a disagreement as evidence of a changing system. Field notes record context that a number alone cannot show, including weather, location, sampling depth, and equipment condition. A well-documented method makes a result easier to evaluate and reproduce.
When a task asks how technology contributes to Earth science, connect the tool to the observation it enables. Satellites support repeated broad coverage; seismometers record waves that reveal subsurface structure; GPS measures surface motion; and remote vehicles reach difficult environments. Each tool answers some questions well and others poorly, so match the instrument to the scale and variable.