Can You Compare Reasoning Practice on Phone and Laptop?

What question can these two sessions actually answer?

Two sessions on a phone and a laptop can tell you which screen feels easier to use and what reasoning operation you want to review. They cannot show which device produced a truer IQ result. Test IQ Free describes its assessment as 50 adaptive questions across pattern reasoning, number patterns, and spatial reasoning, with domain feedback. Its displayed estimated IQ range is provisional: it comes from placeholder conversion parameters and has no validated adult reference population. The product description provides no evidence that its scores are interchangeable across devices or a rule for correcting one device’s result.

That distinction leaves a useful personal question. You can describe your experience on each screen—whether a diagram was easy to inspect, whether controls felt comfortable, or whether you could concentrate—and choose a worked answer or operation to revisit. Those are observations about access and practice. They do not establish a device effect on ability, a trend in IQ, or which estimate should be preferred. The number may look like a compact summary, but its apparent precision does not supply the missing evidence for that interpretation.

The measurement research discussed later can help explain why device comparisons sometimes differ and why a second sitting brings another change into the picture. Those studies concern other tasks and assessments; they can frame sensible questions, but they do not directly test this adaptive form or validate its conversion. So the useful output from two visits is a modest record of what worked for you and what to review, not a calibrated comparison of intelligence across screens.

For example, a hypothetical note might say, “The diagram was easier to inspect on the laptop.” That reports a concrete experience. “The laptop measured my ability more accurately” makes a claim about measurement quality that the experience alone cannot support. Keeping those statements separate lets the first observation guide a practical choice without turning it into a score conclusion. You need not treat the test as useless: it offers challenging reasoning practice and worked answers, while the range itself remains a provisional estimate rather than an established comparison with adults.

A sensible comparison therefore starts with a narrow prompt: which setup made it easier to see the material and stay with the task, and which explanation do I want to revisit? If you make a note, include the screen and the order of the visits so that you remember the circumstances. Avoid treating the second displayed range as a correction of the first, averaging the two, or selecting the higher one as the better estimate. None of those operations has a supported calibration here. The comparison is worthwhile when it helps you choose a usable screen or a next practice target; it need not become a verdict about your capacity.

Because the assessment adapts difficulty, a second visit should not be assumed to reproduce an identical item path merely because it has the same question count. That is another reason to describe the experience rather than treating the displayed ranges as if they came from a controlled device trial. You can still notice whether the screen helped you follow a diagram or whether a particular worked answer clarified a pattern. Such a note is useful precisely because it makes a limited claim you can act on without requiring a score comparison.

Sources: Free IQ Assessment: 50 Adaptive Questions, Estimated IQ Range | Test IQ Free

What has to be comparable before a score comparison means anything?

Matching the name of a test across two screens is not enough to make its scores comparable. A defensible interpretation depends on what the assessment measures, how it is administered and scored, who the evidence concerns, and what decision the score is meant to support. The Standards for Educational and Psychological Testing frame validity as evidence and reasoning that support a proposed interpretation and use of scores. Validity is therefore not a permanent badge attached to a test title: evidence for one population, administration, or purpose does not automatically answer a different question.

Delivery mode is one condition that may matter to the interpretation. A phone and a laptop can present different screen sizes and input methods; whether those differences change performance depends on the specific tasks and how responses contribute to the score. To claim that scores can be treated alike across modes, evidence needs to address the relevant assessment, scoring, population, and intended use. A shared label identifies neither identical task demands nor equivalent measurement. Conversely, a difference in screens does not by itself prove that one mode distorts a score. The question calls for evidence matched to the claim.

Researchers sometimes examine this question with measurement invariance: broadly, whether the relationship between the measured construct and observed responses behaves comparably across groups or conditions under a specified statistical model. The tested level matters, as do the groups, instrument, and conditions in that analysis. Evidence that a designed assessment meets invariance criteria across smartphone and desktop groups is informative for that assessment and those tested comparisons. It does not certify every digital test, every device, or a one-person comparison on a different assessment.

Group-level evidence and paired personal observations answer different questions. A study can compare distributions or model relationships across samples under defined conditions; one person's two sittings show what happened on those occasions, with device, order, and prior exposure entangled. A reader does not need to run a formal invariance analysis to use a screen that feels comfortable. The framework instead sets a boundary on what to conclude: experience can guide usability and review, while score interchangeability requires evidence that actually covers the proposed interpretation. Later examples can then be read as conditional findings, not universal device rules.

Several distinct ingredients sit inside the word “comparable.” The construct is the ability or operation the score is intended to represent; administration includes the conditions and interface through which the person encounters tasks; scoring turns responses into a result; and the reference population determines what a comparison group can mean. A score might be consistent under one set of conditions yet still lack evidence for another interpretation. Likewise, evidence that a mode changes completion time would not, on its own, establish a change in accuracy or in a composite score. The outcome under discussion must be named rather than assumed from the broad label “performance.”

This is why a validity argument is tied to a use. Evidence for describing performance in a research task is not automatically evidence for ranking an individual, making an educational placement, or comparing an online estimate with a professional instrument. Each step adds a claim about what the score means and what action it can support. The Standards framework asks readers and test users to attend to the population and administration conditions relevant to that interpretation; it does not make a test valid simply because it follows a familiar format or is delivered consistently.

For a personal phone–laptop comparison, the practical test is therefore not whether the screens look similar, but whether evidence exists for the same score interpretation under both conditions. If it does not, an observed difference remains a difference between two sessions. It may prompt a question about comfort or task demands, but the score alone cannot identify its cause or settle which result better represents the person.

Sources: Standards for Educational and Psychological Testing: Open Access Files

What have personal-device studies found, and where do they stop?

The large personal-device study, “Cognitive test scores vary with choice of personal digital device,” examined visitors who chose their own device on TestMyBrain.org. Its first sample included 59,587 visitors and its second 3,818. Across the two samples, the researchers compared performance on online cognitive tasks delivered through participants’ phones, tablets, laptops or desktop computers. This is useful evidence that device choice can accompany measurable differences in particular online tasks at scale. It is not a randomized trial in which otherwise comparable people were assigned a phone or computer.

The outcome matters. Among tasks that measured reaction time, mobile-device users tended to respond more slowly than laptop and desktop users; the pattern was most pronounced for Android smartphone users. In the second sample, the researchers adjusted comparisons for age, gender, education and performance on an untimed vocabulary task. The slower mobile reaction-time pattern remained after those adjustments. Adjustment makes the measured comparison more informative than a raw split between device groups, but it does not turn the groups into randomly assigned equivalents or remove every possible difference between people who chose different devices.

The study also reported that device-interface groups differed on some tasks whose demands were not identical. The pattern depended on what the task asked participants to do: differences appeared on tasks requiring fast responses and on tasks involving fine motor movement. The paper described links between touchscreen use and simple reaction-time performance, and between screen size and performance on trail-making tasks. Those findings do not amount to a single ranking in which one device wins every measure. A fast response, a precise touch, and navigating a task across a display are different requirements, so an association on one outcome cannot be silently translated into a claim about all cognitive performance.

The untimed vocabulary measure adds a useful complication. Device groups also differed on that measure, even though it did not depend on speeded responses in the same way. This makes it plausible that who chose a device contributed to some observed group differences, alongside any influence of the interface. People may select a phone or computer for reasons tied to their circumstances or habits; the study’s adjustments address specified measured characteristics, not every factor that could separate those groups. Since participants used their own equipment, the researchers observed a bundle of device, user and testing-context differences rather than isolating hardware by experimental assignment.

The two-sample design also matters to how broadly to read the pattern. The larger visitor sample gives the researchers a substantial set of observations, while the second sample allows them to examine the relationship with additional measured characteristics, including vocabulary. Neither sample represents a controlled comparison of the same people using different devices under otherwise matched conditions. The measures were tasks available through that site, not a uniform test battery that can be assumed to behave like every reasoning quiz. Large numbers can make an association easier to detect; they do not by themselves identify its cause or make it portable to a new instrument.

That distinction sets the boundary on the result. The study supports the narrow statement that, among these TestMyBrain visitors and tasks, the device people used was associated with differences in some measured outcomes, including slower mobile reaction times after the reported adjustments. It cannot establish that using a phone caused a performance penalty, that laptops are generally superior, or that every device difference was adverse. Nor does an observed association tell us how much of a difference belongs to screen or input characteristics and how much to the people or settings represented in each group.

It also cannot supply a correction for a different assessment. TestMyBrain’s tasks and outcomes are not the same as Test IQ Free’s adaptive 50-question set of pattern, number and spatial reasoning questions. No result in this BYOD comparison validates a phone-to-laptop adjustment for this test, says how a device difference should enter its scoring, or establishes that its provisional estimated IQ range has the same meaning on each device. The range uses placeholder conversion parameters and has no validated adult reference population; these personal-device studies do not fill that evidence gap. Their practical contribution is more limited and still useful: device conditions can matter for particular online tasks, but the direction and meaning of a difference need evidence about the task and assessment at hand.

The study’s value is therefore comparative rather than prescriptive: it identifies outcomes worth checking in future work, not a device rule for individual users. A person’s experience may motivate a usability choice, but it cannot substitute for that assessment-specific evidence.

Sources: Cognitive test scores vary with choice of personal digital device

How can hardware timing and task design change what is observed?

Device latency is the elapsed time added between software requesting a visual stimulus and that stimulus physically appearing, or between a person’s input and the software registering it. In a phone, for example, part of that interval can occur in the display and part in the touchscreen response. The study “To BYOD or not: Are device latencies important for bring-your-own-device (BYOD) smartphone cognitive testing?” measured this timing with a robot on 26 Android and iOS smartphones. Across those phones, the combined display-plus-touch latency ranged from 35 to 140 milliseconds.

That range documents variation in hardware timing; it does not report how people reasoned or how many answers they got right. The robot registered when the device displayed a requested event and when it registered a touch. No reasoning-test participants took part in this measurement, and the study did not compare IQ scores. Its result therefore identifies a technical condition that can be relevant to measurement when a task records response speed, rather than a demonstrated score shift for any particular reasoning assessment.

Timing is most directly relevant when the recorded outcome is itself a response time or when an answer must be made within a tightly controlled interval. If a task measures how quickly someone taps after a signal, delay in displaying the signal or registering the tap can become part of the recorded interval. A difference of tens of milliseconds may then matter to the precision of that timing measure, depending on the task and measurement setup. The robot study’s range makes the possibility concrete, but its measurements alone do not quantify the effect for every task or device pair.

The implication is different when the outcome is whether a person selects a correct answer without a response-speed score. Extra time between a touch and its registration could affect the experience of interacting with an interface, but latency by itself does not show that the reasoning behind a particular answer was wrong. Nor can a hardware latency value be converted into an IQ adjustment: that would require evidence about the assessment’s design, response handling and scoring. The product profile does not establish whether response timing enters Test IQ Free’s provisional estimated range, so it would be speculation to say that this measured phone latency raises or lowers that range.

Task demands help explain why the observational patterns in “Cognitive test scores vary with choice of personal digital device” were not uniform. A task asking for a quick reaction places different demands on timing than one requiring small, controlled movements or work across a larger visible area. Screen size and input mode may align differently with those demands: a compact touchscreen may be relevant to a rapid tap, while display area may matter to navigating a task that occupies more of the screen. These are plausible ways to interpret the study’s task-specific associations, not experimentally established mechanisms for every assessment. Latency is only one possible influence; user characteristics and setting remain part of a BYOD comparison.

This distinction also prevents a common interpretive shortcut: treating any slower response as weaker reasoning. A timing measure records elapsed time under a particular interface and task rule. Whether that time is part of the construct of interest, a nuisance in its measurement, or irrelevant to a scored answer depends on what the assessment actually records. The hardware study establishes that devices can differ in timing; it does not decide which interpretation applies to a separate test.

Consider a simple explanatory contrast, not a measurement of this product: timing a button press after a cue is more exposed to input delay than judging whether an untimed diagram is legible. In the first case, the recorded time includes the path from action to registration; in the second, screen clarity and comfort may affect the experience without creating a response-time value. A larger screen could make a diagram easier to inspect while leaving reasoning accuracy unchanged. This example separates timing from visibility, user-interface comfort and scoring; it is not evidence that any one of them changed a Test IQ Free result.

The practical conclusion is conditional. Hardware timing gives researchers a reason to examine devices carefully when speed is part of the measured outcome. It does not show that all phone-based reasoning is less accurate, that a computer is always better, or that this adaptive test applies a timing penalty. To make any of those claims would require a study of the relevant test, outcome and scoring process. The latency measurements explain one possible source of variation; they do not settle what a person’s answer means.

Sources: To BYOD or not: Are device latencies important for bring-your-own-device (BYOD) smartphone cognitive testing?; Cognitive test scores vary with choice of personal digital device

Why can one device comparison disagree with another?

Device effects are not inevitable, and the studies do not support a universal rule that one screen produces the better score. In “Comparison of PC and iPad Administrations of the Cogstate Brief Battery in the Mayo Clinic Study of Aging,” researchers examined cognitively unimpaired adults aged 50 and older across three cohorts. One cohort of 168 people completed matched same-day PC and iPad administrations; separate baseline PC and iPad cohorts included 1,820 and 605 participants; longitudinal groups included 63 people in each mode. Those designs let the investigators examine both same-day mode differences and patterns across administration groups, but the measures and participants were specific to the Cogstate Brief Battery and the Mayo Clinic Study of Aging.

Results depended on the outcome. Several Cogstate measures were completed faster on PC, while accuracy generally showed little difference between modes. Groton Maze outcomes went in another direction, so the pattern was not simply “PC better” across the battery. The study described the mode differences as small but statistically significant and called for further work on their clinical meaning. That qualification matters: a statistically detectable difference does not by itself tell a clinician whether the difference changes an interpretation or decision. Nor does the combination of faster completion and similar accuracy establish that every task is interchangeable; speed and accuracy answer different questions, and the Groton Maze pattern resists reducing them to one device ranking.

The three-cohort structure also matters when reading the Cogstate result. The same-day matched cohort places each participant in both modes within a short period, so the comparison addresses mode under paired conditions for that group. The baseline PC and iPad cohorts, by contrast, are separate groups; their mode contrast can also reflect differences between the people who entered each cohort. The longitudinal groups add repeated measurement over time, where mode and occasion are both relevant. These designs provide complementary evidence, but none should be described as a single randomized experiment in which every participant was assigned to an otherwise identical device condition. The reported outcome pattern must be read alongside which cohort supplied it.

The second study provides a genuine counterexample to the idea that mode differences must occur. “Equivalent Selective Attention Scores From Different Digital Devices: On the Fairness of Assessments Designed for Smartphones, Tablets, and Desktop Computers” reported a 2025 mobile-first selective-attention assessment tested with 296 matched participants using desktop or smartphone. The researchers used confirmatory multigroup factor analysis and reported measurement invariance at all investigated levels. For that assessment and those tested groups, the result supports the claim that its measurement relationships were comparable across the modes under the reported analysis. It shows that device comparability can be designed for and examined, rather than assumed to be impossible.

The details also show why “performance” needs an outcome attached. Completion speed, accuracy, and performance on a maze task are not interchangeable labels for one underlying device effect. A faster response can coexist with little accuracy difference, while another task can show a pattern in the opposite direction. Even within one battery, a summary such as “PC participants did better” would conceal which measure changed and whether the difference mattered for the intended interpretation. The authors’ call for more work on clinical meaning leaves that last question open: statistical significance describes evidence against a no-difference model under the analysis, but does not by itself quantify practical importance for a particular decision.

These findings can coexist because the instruments, operations, design choices, participants, and outcomes differ. The Cogstate comparison found small, outcome-specific mode differences in a particular battery and older adult sample; the mobile-first study found invariance for a different assessment built around selective attention. Neither finding cancels the other, and averaging them into a general device effect would erase the very conditions that explain why they diverge. One assesses a set of cognitive battery outcomes, the other a defined selective-attention construct, with a design deliberately intended for multiple devices.

The practical inference is that device comparability is an empirical question about a particular assessment and interpretation. A positive invariance result is meaningful evidence for the assessment tested, while a mode difference on a different battery is evidence about its own outcomes and conditions. Neither determines whether this adaptive reasoning test behaves equivalently on a phone and laptop. Its own mode evidence would be needed to answer that question; the studies here demonstrate why conclusions should remain tied to the assessment that generated them.

The mobile-first study asks a different but equally bounded question. Because the assessment was designed for use across devices, its invariance analysis tests whether the measurement model remains comparable across the specified desktop and smartphone groups, rather than merely asking whether raw averages happen to look similar. That is why the positive result is informative: it supports a defined form of comparability under its analytic criteria. The result is not a guarantee that every score, interface, or user experience is identical. Invariance at the reported model levels concerns the modeled relationships in those groups; it does not erase possible differences in comfort or establish validity for unrelated uses.

Sources: Comparison of PC and iPad Administrations of the Cogstate Brief Battery in the Mayo Clinic Study of Aging; Equivalent Selective Attention Scores From Different Digital Devices: On the Fairness of Assessments Designed for Smartphones, Tablets, and Desktop Computers

How does a repeat sitting add a second source of change?

A second sitting changes more than the screen if the person has already seen the assessment. “Retest effects in cognitive ability tests: A meta-analysis” synthesized longitudinal evidence from 122 studies, 174 samples, 786 outcomes, and 153,185 participants. The authors used a longitudinal multilevel synthesis and considered up to four administrations, allowing the analysis to examine how score patterns changed with repeated exposure across a broad set of cognitive ability tests. Its outcome was change in test scores after administrations were repeated; it was not a direct test of whether general intelligence changed, nor a device comparison for one specific browser assessment.

The meta-analysis found significant retest effects: scores tended to rise with repeated administration in the pooled results, with pooled gains leveling after the third administration. The authors also reported moderation by cognitive operation or content, form equivalence, the interval between administrations, and age. In other words, the pattern was not one fixed gain applying identically to every test and person. What is asked, whether later forms are equivalent, how long separates sessions, and the ages represented all help condition the observed score changes. The leveling result describes the pooled trajectory; it does not mean every individual follows that trajectory or that a fourth sitting cannot differ from a third.

For someone comparing a phone visit with a later laptop visit, this evidence supplies a competing explanation for a changed result. The later session differs in device, but also in exposure: the person may recognize a task format, remember an approach, or have worked out a strategy after the first encounter. Those are plausible ways prior exposure can matter, not mechanisms that this meta-analysis isolates for any one person. Since the review combines heterogeneous tests, its pooled finding cannot tell a reader how much a particular second result changed because of familiarity, a device, another session condition, or chance variation.

The review’s scale gives a broad view of retesting, but breadth also sets a boundary. Its 786 outcomes are drawn from many instruments and administrations, not 786 interchangeable versions of one assessment. A pooled pattern can reveal that repeat exposure is a recurring measurement consideration across the literature while leaving the size and direction for a specific test unresolved. The reported moderators reinforce this point: differences in cognitive operation or content, how equivalent the forms are, the interval, and participants’ ages help shape the observed retest pattern. Without matching those features, a reader cannot lift a pooled result as a personalized adjustment.

Repeated exposure and targeted practice are related but distinct ideas. Retaking an assessment brings another administration of its tasks or forms; deliberate practice may instead focus on an operation, explanation, or fresh example. The meta-analysis establishes that repeated administration is a relevant source of score change in the studied cognitive ability tests. It does not show that any change reflects broad intellectual growth, and it does not estimate the effect of practicing the operations in this site's test. Nearer familiarity with a test and transfer to unrelated abilities are different claims, which require different evidence.

This distinction changes how to describe an apparent improvement. If a later result is higher, it is accurate to say that the later sitting produced a higher result in that sequence. Calling it a device benefit would require separating the screen from the prior encounter; calling it a general ability gain would require evidence that the change reflects a broader construct rather than familiarity or task-specific learning. The meta-analysis does not make either separation for a particular pair of visits. Its contribution is to show why administration history belongs among the possible explanations, not to choose an explanation after the fact.

The fair use of the pooled result is therefore modest: remember that prior exposure can accompany score change and include it in the description of a personal sequence. It is a reason not to attribute the entire difference to the phone-to-laptop switch. It is not a forecast that this reader's second score will rise, a diagnosis of why it changed, or a reason to avoid low-stakes practice. Some people may show little change or a different pattern; the aggregate analysis does not identify which case applies to an individual.

Because the reviewed studies vary in content, form, interval, and participant age, the meta-analysis offers no single correction or waiting period for this reader. Applying its pooled pattern to the site's adaptive 50-question assessment would require test-specific evidence that the review does not provide. A useful personal record can simply preserve the order and note whether this was a first or later exposure. That small detail helps keep the device observation in context, without turning a group-level result into an explanation or prediction for one person's estimated range.

Sources: Retest effects in cognitive ability tests: A meta-analysis

What kind of progress can practice notes honestly capture?

Practice notes can capture whether the test and its explanations work for you in practical terms, and which specific reasoning operation you want to revisit. They can record what you noticed while using the interface or studying a worked answer. That is genuine information about the experience and your learning process. It does not turn the provisional estimated IQ range into a sound measure of progress, and it does not establish that general intelligence changed.

A useful note distinguishes an observation from an interpretation. ‘The diagram was hard to inspect on the small screen’ records an experience. ‘The phone lowered my reasoning ability’ assigns a cause and meaning that the observation alone cannot show. Likewise, ‘I followed the explanation of the rotation step’ describes a particular learning outcome; it does not show that the same skill improved across unfamiliar tasks. Keeping these statements separate lets a practice record remain useful without giving it more evidential weight than it has.

For example, a hypothetical note might read: ‘Phone first; laptop second; the rotation explanation was easier to inspect on laptop; revisit the worked answer.’ This is an illustration, not a report of an actual participant or test result. It preserves the order, identifies a concrete difference in viewing the material, and names a next step. It contains no score comparison and does not claim that the laptop caused better reasoning. A reader could later use the note to decide which screen makes diagrams more comfortable to examine, while still leaving the source of any numerical difference unresolved.

The same kind of record can help identify usability conditions. A person might note that diagrams were legible, that controls were comfortable to use, that an interruption broke concentration, or that a worked explanation made one operation easier to inspect. These observations answer practical questions: Could I see the task? Could I interact with it comfortably? Was the setting suitable for concentrating? Did the explanation clarify the step I wanted to understand? They can guide a choice about where to take an open-practice test or how to review its answers. They do not establish a device effect on ability or a corrected score.

A note can also make practice more specific. Rather than writing ‘get better at reasoning,’ name the operation that prompted a question: tracking how a shape changes orientation, for instance, or checking how a numerical sequence changes from one term to the next. Then describe the step in the worked answer that you want to understand. This turns a broad intention into a review target. It does not imply that the assessment has measured the operation precisely enough to diagnose a weakness; it simply records what you chose to study after seeing an example.

Learning a method or becoming familiar with a task format is a legitimate, narrow outcome of practice. If an explanation makes a transformation easier to follow, or a person becomes more comfortable with the way a sequence problem is presented, that can matter for the next encounter with similar material. But familiarity with a format is not proof that general intelligence rose. Nor does success on a practiced operation establish a gain on unrelated, unfamiliar tasks. Those are broader claims than a personal note or a repeated browser result can support, and this section does not assume that any such transfer has occurred.

This boundary does not make practice pointless. It clarifies what a reader can reasonably aim to learn from it: understand a particular worked step, notice whether the display supports inspection, or become more deliberate about an operation. The free test provides worked answers for review; a reader can use them to revisit the reasoning shown, then try fresh practice if desired. A useful record would say what operation to revisit and what about the explanation or interface mattered. It need not attach an IQ label, calculate a personal gain, or promise a change in general ability.

A numerical result is attractive because it compresses an experience into one comparison. In this setting, that simplicity exceeds what the provisional estimate and a personal practice note justify. The range comes from an unvalidated conversion without a validated adult reference population, so it is not an established population comparison or a validated tracking scale. A written observation can still support a proportionate next action: review an answer, choose a more comfortable screen, or explore a fresh example of the named operation. The useful progress is the specific understanding or usability information the reader can describe, not an inferred rise in general intelligence.

Why does session order make a two-device story hard to interpret?

A phone-first, laptop-second sequence cannot isolate a device effect because device changes at the same time as session order. The first sitting is also the first encounter; the second comes after the person has already seen the assessment. Time has passed, and the conditions of the two sessions may differ. A result difference can therefore be described as occurring between those two sittings, but the sequence alone cannot show which changing condition produced it.

Imagine a hypothetical person who completes the assessment on a phone and later completes it on a laptop. In the second sitting, the device is different, but so is the person's position in the sequence: they are no longer encountering the test for the first time. They may recognize how an item is presented or remember a prior approach. The time of day, surrounding quiet, fatigue, interruptions, and other session conditions may also differ. These are possible differences to consider, not claims that any one of them affected this hypothetical person's answers.

The missing counterfactual is the same person's first encounter, at the same moment and under the same conditions, on the other device. After taking the test on the phone, that exact first encounter cannot be repeated on the laptop: the person has already encountered the material and the passage of time cannot be undone. If the order were reversed, the laptop would be first and the phone second. That would change which mode is paired with the earlier and later session, but a single person still could not experience both devices as the first device in the same encounter.

This is why changing the order in one's own two sittings does not fully solve the comparison. If phone comes second, it is now the later session; if laptop comes second, it is the later session instead. Device and order remain linked within each two-sitting story. A reader can preserve the sequence accurately, but cannot use that record to calculate what the second result would have been had the first device been used again under identical conditions. The absent alternative is not a missing opinion; it is an observation the person cannot obtain after the first sitting has occurred.

Researchers can address order more systematically through counterbalancing. In plain terms, different participants are assigned different orders: some encounter mode A before mode B, while others encounter B before A. Across a sample, researchers can compare patterns associated with order and mode instead of allowing every participant to experience the modes in only one sequence. The method helps estimate whether observed group patterns are consistent with a mode difference, an order difference, or both, given the study's design and analysis. It is a way to improve a group comparison, not a label for what one individual should do.

Counterbalancing does not make every source of variation disappear. Participants can differ, and conditions still need to be considered. The point is that varying the sequence across people gives a study information that one fixed-order comparison lacks. It cannot retroactively balance one reader's completed sittings or reconstruct that person's unobserved first laptop encounter. A personal phone-first/laptop-second pair therefore remains descriptive, even if the reader keeps careful notes about dates, settings, and what was used. It can tell what happened in that order; it cannot supply the other order for the same first encounter.

There is still practical value in a within-person comparison. A reader can notice that one screen made diagrams easier to inspect, or that controls felt more comfortable in a particular setting. That experience can inform a usability choice for future open practice. Causal ambiguity limits the interpretation of a numerical difference, not the reality of the experience the reader recalls. Describing the observation plainly keeps both points intact: the reader noticed something meaningful about using the assessment, while the two-session design does not identify why a score or range differed.

The responsible account is therefore specific about order: ‘I took it on my phone first and on my laptop second, and the results differed.’ Calling the difference a laptop effect would attribute it to one factor while device, prior exposure, time, and session conditions changed together. The test's displayed range is itself provisional, based on an unvalidated conversion with no validated adult reference population; that further limits its use as a comparison target, but it does not alter the basic order problem. Preserve the sequence in any note and describe the numerical difference without assigning it to the device.

What should you record and what should you do next?

Keep a record only if it helps you choose a usable setup or a worthwhile review target. For one session, choose the screen that feels comfortable and start there. If you deliberately use both, write which came first. That simple order note describes the sequence; it does not convert the two visits into a controlled comparison. You do not need to repeat the test or keep formal notes to get value from open practice: one comfortable sitting and a review of an answer may be enough.

If you are deciding whether to use a phone or laptop, base that choice on the task in front of you: can you see the diagram and comfortably use the controls? Pick the screen that lets you attend to the reasoning. You do not need to test every device, adjust brightness by a formula, or standardize a personal setup. A note such as ‘phone; diagram visible; no interruption’ can be enough to remember the conditions that mattered to you. It documents access, not a score-quality rating.

After a sitting, add one observable access or process fact. For example, note that a diagram was easier to inspect on the larger display, or that an interruption broke your concentration. These are observations about what happened during the session. Avoid turning them into a cause for a score difference: ‘I could inspect the diagram more comfortably’ reports an experience, while ‘the laptop measured my ability better’ makes a measurement claim the observation cannot establish. Keep any displayed result separate from this short description.

If nothing about access or the explanation stood out, leave that part blank rather than filling a template for its own sake. The purpose is to retain one detail that may help with the next practical choice, not to build a complete account of every condition. A brief note can also remind you that a later visit followed an earlier one, without requiring you to explain any difference between the results. Keep the record private and simple enough that making it does not interrupt the practice you wanted to do.

Then name one operation or explanation to revisit, such as following a shape rotation or tracing how a number sequence changes. A useful note might read: ‘Phone first; laptop second; the rotation explanation was easier to inspect on laptop; review that worked answer.’ This is an illustrative format, not a reported result. It preserves the order, records one concrete access fact and gives you a next step without calculating a personal gain. Stop keeping the note once it no longer helps you choose a screen or focus your review.

A small record works best when it stays concrete. ‘The sequence explanation helped me see that each step alternates’ identifies something to review; ‘my number reasoning is weak’ turns one encounter into a broad label. You can return to the worked answer and then try a new example that uses the same operation. The fresh example lets you check whether you can follow that reasoning again in a different presentation, but success still describes that practice, not a validated change in general ability. Keep the language close to what you saw and did.

Do not average ranges, choose a ‘best’ result, or apply a numerical correction. Test IQ Free describes its range as provisional, generated from placeholder conversion parameters without a validated adult reference population; the product provides no calibration for those personal calculations. Its page, ‘Free IQ Assessment: 50 Adaptive Questions, Estimated IQ Range,’ also says unfinished progress saves locally. It does not establish cross-device synchronization, so if progress is unfinished, treat switching devices as a new visit unless the live product explicitly confirms otherwise.

The record need not travel with the score or become a running log. If your only question is whether a screen is comfortable for a one-time practice session, make that choice and move on. If you return to an answer later, the operation you named is enough to find your place. This keeps the note tied to an action you can take, rather than encouraging repeated sittings just to produce more numbers.

For low-stakes practice, take the free test at [/test](/test), use its worked answers, and try a fresh example of the operation you chose. Understanding an explanation or becoming more familiar with a format is a specific practice outcome; it does not show that general intelligence has increased. The optional $9 report can add explanation and practice structure, but it uses the same provisional conversion and adds no validity. You can skip both the second sitting and the notes if a single session already answered what you wanted to know.

Sources: Free IQ Assessment: 50 Adaptive Questions, Estimated IQ Range | Test IQ Free

When should a browser result stay out of a consequential decision?

For curiosity and open practice, the free test is a reasonable place to start; treat its displayed range as provisional. If the result may affect education, employment, diagnosis or eligibility, ask a qualified professional what assessment fits that specific decision and population. The *Standards for Educational and Psychological Testing* frame validity around evidence for a proposed interpretation and use, while APA’s *Ethical Principles of Psychologists and Code of Conduct, Standard 9.02: Use of Assessments* directs psychologists to consider suitability and explain limitations. Those standards do not endorse this browser estimate for consequential interpretation, and they do not mean a casual user needs clinical testing. A qualified professional can clarify whether a formal assessment is needed at all, which population and purpose the available evidence covers, and what the resulting score can and cannot answer. This advice is especially relevant when an institution or decision-maker needs evidence, rather than when you simply want to try reasoning questions. You can bring relevant context and ask directly instead of relying on a recreational result to answer the decision. A useful question to take to that conversation is: “I need evidence for this decision; what assessment, if any, fits, and what limits should I understand?”

Sources: Standards for Educational and Psychological Testing: Open Access Files; Ethical Principles of Psychologists and Code of Conduct, Standard 9.02: Use of Assessments

Questions readers ask