Average IQ in Canada: how norms change the meaning of a score

0 The short answer

Canadian normed IQ scales use a reference frame centered near 100, but that convention is not a new survey of the current Canadian population. Canada's 99.62 in a historical compilation and 102.09 in an online participant table come from different samples and calculations. A published comparison that scored the same Canadian participants using US norms illustrates why the norm group matters. This page examines those sources and keeps provincial education results separate from IQ. A country average, whatever its quality, cannot establish an individual Canadian's cognitive score.

1 Four Numbers, Four Frames

Every figure quoted for the average IQ in Canada belongs to a frame, and the table below states the frame before the number, because the number is meaningless without it. An IQ is a position within a norm sample, expressed on a scale with a mean of 100 and a standard deviation of 15, as the page on the standard deviation of 15 sets out. Change the norm sample and the same raw performance receives a different number. The Canadian comparison illustrates that rule: the same standardization participants were scored under two norm systems, and the published Full Scale comparison concerns those with complete Full Scale data.

The table should be read as a set of different questions. A page quoting 104.51 as the present IQ of Canada would generalize a historical norm sample comparison into an unmeasured current population claim. Calling the country exactly 100 without qualification makes a different error: the reference center is set by the norming procedure, while a new sample can have a different mean. The norming guide explains that distinction. A hypothetical individual score of 104.5 on a normal SD 15 scale corresponds to about the 62nd percentile, our arithmetic. That calculation does not place the entire Canadian population at a single percentile or establish the position of an individual Canadian.

2 Why Canadian Norms Center Near 100

Several Wechsler editions have Canadian norms that center their reference scales near 100; a current national mean remains an empirical question. Miller and colleagues, writing in 2015, date the practice to the WISC-III Canadian standardization study of 1996, undertaken because Canadian practitioners had argued since the 1980s that American norms did not represent Canadian performance. The WAIS-III received a Canadian technical manual in 2001, and Lange and colleagues describe its standardization sample of 1,105 adults in Archives of Clinical Neuropsychology. The WISC-IV received Canadian norms in 2004. The WAIS-IV CDN followed in 2008, for ages 16 years 0 months to 90 years 11 months, with core subtests taking 60 to 90 minutes, a qualification level of C and, on the product page read on September 20, 2026, a basic kit price of 3,014.90 Canadian dollars, which will change. A francophone version, the WAIS-IV CDN-F, published in 2010 for the same ages and labeled a research edition, carries Canadian norms drawn from what the page calls a fully stratified Canadian sample. The WISC-V CDN was published in 2014 for children aged 6 years 0 months to 16 years 11 months.

The size and design of the adult sample are documented in the Canadian manual and in the papers that used it. Dutton and Lynn, citing page 42 of the manual in Open Differential Psychology, describe a standardization sample of 688 people aged 16 to 90, representative of the population by sex, age, education, ethnicity and geographic region on the 2006 Census, drawn from all the major regions except the three sparsely populated territories, with parental education used for participants aged 16 to 19. Of the 688, 681 completed the subtests required for a Full Scale IQ, because two subtests were not given to the oldest participants. Miller and colleagues add that Canadian norm samples typically contain 75 to 100 people per age band, and in some bands as few as 30, against roughly two to four times as many in the American bands, and discuss the implications of sample composition for the stratification design. That argument should be evaluated as a methodological claim, not a statement about fixed traits of Canadians. Pearson's product page says only that the norms rest on "a fully stratified representative Canadian sample" and does not print the count.

The comparison with the UK evidence is useful only after specifying the edition and reference frame. The British WAIS-IV retained US norm tables after a validation study, while the Canadian edition provides Canadian tables. A score therefore refers to a different reference frame depending on the edition used. Centering the Canadian reference scale near 100 does not prevent the mean of a later Canadian sample from changing on those fixed tables. The target population can change, the sample can differ by chance or coverage, and administration conditions can differ. Those empirical possibilities remain even when the score conversion itself is unchanged.

That next edition is not yet available on Canadian norms. Pearson Canada's page for the WAIS-5 US version, read on September 20, 2026, lists a 2024 publication date and "US Norms", states that "WAIS-5 Canadian is currently in standardization", and names three situations in which American norms are appropriate: alongside other American normed tests, for preparation, and for a client who has spent most of their life in the United States. The company's standardization study page says the Canadian study was "expected to begin in Summer 2025 and continue until early 2026" and recruits residents of Canada aged 16 to 90. Within the adult Wechsler products reviewed on that date, the WAIS-IV CDN supplied Canadian norms while the listed WAIS-5 US edition used American norms. This is not an exhaustive inventory of every assessment available to Canadian adults. The page on what the WAIS-5 is describes the American edition, and the page comparing the WAIS-4 and WAIS-5 sets out what changed between editions.

3 The Same Canadians on American Norms: 104.5

The most informative Canadian number is not a national IQ but a comparison of two norm samples on one instrument, and it shows the Canadian sample scoring higher and spreading less than the American one. Miller, Weiss, Beal, Saklofske, Zhu and Holdnack report that across all age groups and most subtests the Canadian WAIS-IV sample obtained higher raw scores than the American sample, that its mean Full Scale IQ on American norms was 104.51 against the American 100, and that its standard deviation was 1.6 points smaller. Table 4.6 of the Canadian manual, which they cite, puts the Canadian mean scaled scores on American norms between 10.1 and 11.0 for the subtests and between 102.3 and 105.0 for the composites. The Canadian raw score distributions were also more negatively skewed, so the two conversions diverge most in the lower tail. The pattern is old. Holmes found in 1981 that 340 children in British Columbia scored a mean Full Scale IQ of 106.95 on the WISC-R, with a standard deviation of 12.66, and the WISC-III Canadian standardization of 1996 found Canadian children 3.34 Full Scale points, 1.94 Verbal IQ points and 4.96 Performance IQ points above the American norms, both according to Miller and colleagues' account of the manuals.

Pearson discusses sample composition in a December 2014 note. It describes Canadian and US norm differences across several editions and discusses the lower variability of the Canadian standardization data. Those observations concern the sampled distributions; they do not establish a fixed national trait or guarantee a lower Canadian score for every person and index. Miller and colleagues report that 56.7 percent of the Canadian WAIS-IV norm sample had attended college for a year or more, compared with 43.9 percent of the US sample. After matching on age, sex, ethnicity and education, differences were reduced or absent, with remaining effect sizes between .11 and .31 and no difference among ages 70 to 90 in that analysis. Dutton and Lynn's matched samples of 488 and 101 participants reduced the reported gap to 2.1 points. Matching changes the comparison and cannot identify all causes of the original difference. Ford and colleagues' Woodcock-Johnson III comparison, described by Miller, also reported different sample spreads, with standard deviations of 13.52 and 15.72. These are study-specific results, not proof that nationality determines either a person's score or its variability.

The gap has been the subject of a published dispute, and a reader should know both sides. Harrison, Armstrong, Harrison, Lange and Iverson, in Archives of Clinical Neuropsychology, rescored 432 Canadian postsecondary students who had been assessed for learning disabilities, attention deficit hyperactivity disorder or other problems. Their mean Full Scale IQ was 98.4 on the American norms and 91.2 on the Canadian, a mean difference of 7.2 points, and only 23.2 percent of the students had Full Scale scores within five points on the two systems. According to Pearson's note, 17 percent of that sample scored 85 or below on American norms against 40 percent on Canadian norms. Harrison, Holmes, Silvestri and Armstrong extended the sample to 861 Ontario students in the Journal of Psychoeducational Assessment and argued that the Canadian norms could overstate impairment. Miller and colleagues answered that more than two thirds of those students carried clinical diagnoses, that American clinical samples matched on age, education and diagnosis and scored on American norms produced almost the same shares below 85, 40.1 and 44.3 percent against Harrison's 39.4 and 48.9 percent, and that 100 random draws of 681 people from the American sample gave a median mean of 100.1, so the Canadian sample size cannot explain a 4.5 point gap. Pearson's guidance is to choose the norm set in advance and to use Canadian norms for a Canadian client unless the comparison is with other American normed tests.

What the 104.5 does not mean is that the average Canadian is a person with an IQ of 104.5. It is the position of a Canadian norm sample within an American norm sample, on an instrument standardized in the United States on 2,200 people in 2006, according to Dutton and Lynn's account of the American manual, and published in Canada in 2008; the Canadian tables center their own reference distribution near 100. Dutton and Lynn convert the figure to what they call a Greenwich IQ of 102.5 by subtracting the two points by which Lynn and Vanhanen placed the United States below Britain, which adds an assumption to a measurement. The page on what IQ measures explains why a score describes assessed performance relative to a reference group, with validity depending on the intended interpretation and use.

4 The Ten Samples Behind Becker's 99.62

Becker's Canadian figure rests on ten rows of very unequal quality, and the rows explain why the file's own columns give Canada anything from 91.46 to 100.87 depending on how they are combined. The national table lists 4,229 people across the ten samples, with a mean age of about 24, and reports four component values: a simple mean of 91.46, the file's NW summary of 98.76 and QNW summary of 98.37, both using additional subgroup reweighting, and a school assessment figure of 100.87 derived from PISA, TIMSS and PIRLS. The combined measure the file recommends, the mean of the last two, is 99.62. Lynn and Vanhanen's earlier books, recorded in the same file, gave Canada 97 in IQ and the Wealth of Nations in 2002 and 100.4 in 2012, and Rindermann's column gives 101.22. Lynn and Becker's 2019 book The Intelligence of Nations drew on an earlier version of the same file.

The rows sort into three kinds. Five are Wechsler samples, 3,233 people in all, four of them standardization samples or subsamples and one a low exposure group from an environmental study; by our arithmetic on the file they average 103.24 before Becker's corrections and 100.88 after, which is the 104.5 story again in a different file. Two are Raven's samples of schoolchildren from 1972 and 1979 scored against British norms, at 91.22 and 96.20. Three are samples of Indigenous participants, 464 people or 11 percent of the total, entered at 68.46, 76.02 and 75.33. Those three rows deserve a plain description rather than a number. Two come from a 1976 paper in Psychology in the Schools on 160 First Nations children and young adults in northwestern Ontario, scored against American norms from 1947 and 1955, which is why Becker's Flynn correction removes about ten points from each. The third is attributed to a 2009 paper in NeuroToxicology on lead exposure and event related potentials in Inuit children from Nunavik; the paper's abstract describes electrophysiological measures in 27 five year olds and 110 eleven year olds, so the provenance of a row of 304 adults scored on Raven's matrices is not clear from the paper itself. Becker's own quality ratings give the two 1976 rows the lowest overall score in the Canadian set. They are small, old, drawn for other purposes and scored on norms from another country and era, they say nothing about the abilities of Indigenous Canadians today, and what they do in the file is pull the simple mean of the ten rows down to 91.46 while the file's NW summary remains near 99 under its additional reweighting. Directly weighting all ten corrected rows by their listed sample sizes gives 96.75, our calculation, rather than the workbook's NW value of 98.76. The Canadian formula in NAT!Q35 first calculates separate means for seven rows and three rows, then combines them using an additional 0.042 coefficient for the latter group and divides by 1.042. That explicit model choice is not the same as a plain participant-weighted average and does not establish an inherent group characteristic. Excluding those three records entirely gives a size-weighted mean of 99.93 across the remaining seven rows, a sensitivity calculation that does not turn the remaining samples into a representative national survey.

The remaining rows are the ones a reader can learn from. The Quebec row comes from Bouchard and colleagues' study in Environmental Health Perspectives of 362 children aged 6 to 13 in communities supplied by groundwater, which found a 2.4 point IQ decrease per tenfold increase in water manganese and a 6.2 point difference between the lowest and highest exposure quintiles; Becker took the IQ of the lowest exposure group from a figure. The WISC-III row comes from the cross-cultural analysis in Culture and Children's Intelligence, and the Raven's rows from a 1972 comparison in Multivariate Behavioral Research and from the 1999 German edition of the Raven's manual. The two WAIS-III rows come from a paper whose purpose was to build regression equations for estimating premorbid ability, not to report a national mean, and the means themselves come from Lynn's working notes on that sample. The comparison countries in the same column put Canada's 99.62 beside Japan at 106.39, Singapore at 106.17, Germany at 100.61, the United Kingdom at 99.66, Australia at 99.20 and the United States at 97.50, by our reading of the file. The page on average IQ in India shows what the same procedure does with samples that are far less representative, and the pages on Germany and Australia in this batch trace the neighbouring rows.

5 The Self-Selected Online Table: 102.09 From 7,685 Takers

The online figure reviewed here is the mean of people who chose to sit one website's test; the table does not establish comparable coverage of the Canadian population. The International IQ Test country table, read on September 20, 2026, is dated January 1, 2026 and pools 1,212,714 people who took the site's test during 2025, on a scale whose overall mean the site fixes at 100 with a standard deviation of 15. Canada is 13th at 102.09 from 7,685 takers, up 0.44 from 101.65 among 7,888 takers the year before. The United Kingdom is 14th at 101.57 from 7,310 takers, the United States 18th at 101.04 from 22,605, Australia 5th at 104.45 from 4,245 and Germany 38th at 99.32 from 34,350. The site notes that in 84.8 percent of countries the average shifted by less than two points between years, which describes the reported year-to-year movement among its participant groups without establishing stable or representative selection. The ACIS page on average IQ by country uses that table because it supplies a documented table, although selection can differ between countries and years, and it prints the warning beside the tool: the mean is 100 by the site's choice and sample sizes must be inspected before any conclusion is drawn.

Read properly, the table supports one narrow statement: Canadian takers of that test score within about a point of British and American takers of the same test, without establishing that the three countries have comparable selection mechanisms. It does not support the sentence usually built on it, that Canadians are cleverer than Americans by one point or than Britons by half a point, because the table does not establish that its 7,685 volunteers represent the Canadian population or that the score interpretation generalizes to all residents. The page on free versus validated IQ tests explains why a participant mean alone does not establish population coverage or validity, and the review of the International IQ Test examines that site on its own terms. When a forum post quotes one of these figures, follow its actual source and check the population and scale before adopting the claim. This review does not estimate how frequently particular numbers circulate online.

6 Canada Versus the United States and the United Kingdom

The Canadian comparisons with the United States and the United Kingdom need to be read within their methods. The Wechsler comparison directly illustrates what changing norm tables can do to the same tested participants. The Canadian WAIS-IV norm sample scores 104.51 on the American norms of the same instrument, and the gap is documented in the cited norm research, with related comparisons across editions and smaller residual differences after demographic matching. The WAIS-IV UK example reviewed here uses American norms following a British validation study, rather than a separate British standardization sample, as the page on average IQ in the UK documents from Pearson's British product page. In Becker's file Canada's 99.62 sits 0.04 below the United Kingdom's 99.66 and 2.12 above the United States' 97.50, and the American figure rests on 57 samples of 62,296 people with a mean age of about 12, most of them children scored against norms of other years. In the online table the three countries lie within 1.05 points of each other. The page on average IQ in the USA traces the American figures to their sources.

The representative comparisons are not IQ comparisons at all. In PISA 2025, whose Canadian results the Council of Ministers of Education published in Measuring Up in September 2026, Canada scored 510 in science, 490 in reading and 485 in mathematics; the United States scored 502, 490 and 463 and the United Kingdom 511, 494 and 488, against OECD averages of 482, 461 and 463. In the Survey of Adult Skills, whose Canadian results Statistics Canada released on December 10, 2024, adults aged 16 to 65 in Canada scored 271 in literacy, 271 in numeracy and 259 in adaptive problem solving against OECD averages of 259, 262 and 250. Those scales have their own means and standard deviations, defined across participating countries, and the correlation between national means on such assessments and national means on intelligence tests is high, 0.85 to 0.86 in Rindermann's analysis in the European Journal of Personality, which is why Becker folds school assessments into his figure. A PISA point is not an IQ point. Overlapping confidence intervals do not by themselves establish whether a pairwise difference is statistically significant; use the report's comparison tests and sampling design before claiming a reliable ordering.

The honest summary is that Canada, the United States and the United Kingdom are three countries whose samples land within a few points of each other on every scale, whose differences are smaller than the standard error of a single individual's score on any of the instruments involved, and whose only well measured difference, the Wechsler one, is a difference between two norm samples that the publisher attributes mainly to education. The page on score versus percentile shows what a four point difference means on the scale: a whole distribution shifted by less than a third of a standard deviation.

7 By Province: What the Reviewed Sources and PISA Establish

The sources reviewed here do not establish a comparable current population IQ for every Canadian province. PISA supplies provincial student achievement estimates, which answer a different question. Pearson stratifies its Canadian norm samples by region but publishes a single national set of tables; Becker's file lists British Columbia, Ontario, Quebec and Nunavik as the origins of particular samples but has no provincial values; and the WAIS-IV sample, as Dutton and Lynn describe it, excludes the three territories altogether. What Canada does have is one of the largest PISA samples of any country, drawn so that each province and both language systems can be reported. CMEC's report on PISA 2025 states that about 24,600 students from close to 900 schools in the ten provinces were tested between April 21 and June 20, 2025, that the assessment was a two hour computer based test in English or French, and that Canada and most provinces, except Prince Edward Island, New Brunswick and Saskatchewan, failed one or more of the PISA technical standards on participation rates, so that their results "should be treated with caution".

The table is a PISA table, and three things stop it being an IQ table. First, the scale is PISA's own: an OECD mean set at 500 with a standard deviation of 100 when the scale was fixed, so that the 46 point gap between Alberta and New Brunswick in science is a little under half a PISA standard deviation, and the report does not establish a validated direct conversion of that gap into IQ points. Second, the population is fifteen year olds in school, sampled by school, and the report itself says that only differences that are statistically significant should be read; the 95 percent confidence interval for Alberta's science score runs from 509 to 532 and for British Columbia's from 498 to 522, which shows overlapping intervals. A conclusion about the difference should use the report's formal pairwise comparison, rather than infer significance or nonsignificance from overlap alone. Third, PISA measures the literacy that schooling produces in three subjects, and a province's score moves with its curriculum, its language systems and its intake. Quebec leads in mathematics at 500 and sits at the Canadian mean in science but below it in reading; Alberta leads in science and reading; Ontario and British Columbia sit within a few points of the Canadian mean; Saskatchewan, Manitoba and the four Atlantic provinces sit below it in all three subjects, by margins that range from 10 to 36 points in science. The ordering is stable across cycles: in 2022, when mathematics was the major domain, CMEC's previous report placed Quebec at 514, Alberta at 504, British Columbia at 496 and Ontario at 495, with Prince Edward Island at 478, Manitoba and Nova Scotia at 470, New Brunswick and Saskatchewan at 468 and Newfoundland and Labrador at 459, against a Canadian 497 and an OECD 472.

The within province distributions say more than the provincial means. In science, 84 percent of Canadian students reached the baseline Level 2 against 74 percent across the OECD, with the provincial range running from 73 percent in New Brunswick to 85 percent in Quebec and Alberta, and 11 percent reached Level 5 or above against 7 percent, from 5 percent in New Brunswick to 14 percent in Alberta. The difference between the 90th and 10th percentiles in science was 258 points for Canada, the same as the OECD average, and ranged from 241 in Quebec to 269 in Alberta. Thirty nine percent of Canadian students reported an immigrant background against 17 percent across the OECD, and where non-immigrant students outscored immigrant students by 44 points on average across the OECD, Canada showed no gap, with second generation students scoring above both other groups. Students in French and English language systems reached Level 2 at nearly the same rate, 83 and 84 percent, while the English systems had a larger share at Levels 5 and 6; in 2022, students in francophone systems averaged 511 in mathematics against 493 in anglophone systems. None of these numbers is an IQ, and none is a norm. The page on average IQ by state shows how the same proxy problem plays out in the American tables, and the page on IQ and academic achievement explains the individual level relationship that provincial rankings gesture at.

8 The Flynn Effect in Canada: Edmonton, Saskatchewan and the Norm Cycle

Two Canadian school studies contributed early evidence of rising test performance. Their results concern specific tasks and cohorts; the later publication of new norm tables does not itself establish the direction or size of population change. Flynn's 1987 paper in Psychological Bulletin on gains in 14 nations included Canada on the strength of two studies. In Edmonton, all 3,596 Grade 3 pupils in the public schools took Raven's Coloured Progressive Matrices and the California Test of Mental Maturity in May 1956, and in May 1977 the 4,688 Grade 3 pupils took the same tests unaltered, a representative 60 percent of them sitting the Raven's. Flynn's Table 6 gives the Edmonton gains as 6.74 points on Raven's by grade and 8.44 by age over the 21 years, and 9.33 and 11.03 points on the CTMM, which he treats as a Stanford-Binet type full scale score. In Saskatchewan, stratified random samples of classrooms in 1958 and 1978 gave Otis gains of 6.98 by grade and 12.55 by age in Grade 4 and 6.95 by age in Grade 7, though the change from the Otis Beta to the Otis-Lennon makes those estimates less secure. Flynn rated the Edmonton gains as verified and remarked that, projected over 30 years, they amounted to 10 to 12 points on Raven's, "the lowest Ravens gain for any nation with strong data"; the CTMM gains projected to 13 to 16 points, which he called typical of Wechsler and Binet data elsewhere.

The general size of the effect is now measured by two meta-analyses. Trahan, Stuebing, Fletcher and Hiscock's meta-analysis of 285 studies and 14,031 people put the mean gain at 2.31 points per decade, and 2.93 per decade for modern Stanford-Binet and Wechsler tests since 1972. Pietschnig and Voracek's century of global IQ gains, across 271 samples, about four million people and 31 countries, put the gains at 0.28 points per year for full scale scores, 0.41 for fluid and 0.21 for crystallized measures, with a slowing in recent decades. The page on the Flynn effect works through what the numbers mean for a score read against an old norm, and the page on matrix reasoning explains why the fluid tests that dominated Flynn's Canadian data are the ones on which gains were largest.

Canadian norming history provides reference dates, not a continuous measurement of population change. The editions discussed here include the WISC-III Canadian standardization of 1996, WAIS-III Canadian manual of 2001, WISC-IV Canadian edition of 2004, WAIS-IV CDN of 2008 and WISC-V CDN of 2014. Publication dates are not necessarily data collection dates, and the tests differ in tasks and age coverage. Multiplying the years since publication by a pooled historical rate would therefore not predict an individual's current score or an edition-to-edition conversion. A new Canadian WAIS edition could provide useful comparison evidence if its manual reports an appropriate study, but it would not be the only possible design for investigating adult change. The evidence assembled here does not establish a comprehensive current Canadian adult IQ trend. Dutton and Lynn's 2014 suggestion of reversal was a conjecture rather than a reported Canadian longitudinal measurement. Separately, the cited PISA report records science falling from 528 in 2015 to 510 in 2025, reading from 520 in 2018 to 490, and mathematics from 497 in 2022 to 485. Those remain student achievement trends, with the report's participation and pandemic-context qualifications. They cannot be converted directly into a decline in every adult cognitive domain.

9 Canadian Adults: Skills, Education and Age

The Canadian adult surveys discussed here measure literacy, numeracy and educational credentials. They add population context but do not establish how much of the historical Wechsler norm difference any one factor caused. The second cycle of the Programme for the International Assessment of Adult Competencies, collected in Canada in 2022 and 2023, scored Canadians aged 16 to 65 at 271 in literacy, 271 in numeracy and 259 in adaptive problem solving, above the averages of 259, 262 and 250 across the 27 participating OECD countries, according to Statistics Canada's release. Nineteen percent of Canadian adults scored in the two lowest literacy levels, below 226, against 27 percent across those countries. British Columbia scored above the Canadian average in all three domains, at 281, 280 and 269, and Alberta in literacy and adaptive problem solving, at 276 and 264; Newfoundland and Labrador and Prince Edward Island were above the OECD average in literacy and adaptive problem solving, and New Brunswick was at the OECD average in numeracy and problem solving. Numeracy rose slightly since the 2012 cycle and literacy was stable. Adults aged 25 to 34 held the highest scores in every domain and those aged 55 to 65 the lowest, which is the pattern the page on average IQ by age describes for fluid measures across the lifespan.

Education is the gradient inside those numbers. Among Canadians aged 25 to 65, mean literacy was 293 for those with a bachelor's degree or higher, 275 for those with postsecondary education below a bachelor's degree and 246 for those with a high school diploma or less, a spread of 47 points on the PIAAC scale between the top and bottom groups. The 2021 Census, in the Statistics Canada release of November 30, 2022 whose figures sit in table 98-10-0384-01, found that 57.5 percent of Canadians aged 25 to 64 held a college or university credential, the highest share in the G7, and that 32.9 percent held a bachelor's degree or higher, fourth in the G7 behind the United Kingdom at 41.3 percent, the United States at 39.5 percent and Japan at 34.2 percent. Ontario led the provinces at 62.8 percent with a credential and 36.8 percent with a degree, and British Columbia stood at 58.1 and 35.0 percent. The Canadian WAIS-IV norm sample also had a larger share with a year or more of college than the American sample, 56.7 versus 43.9 percent. Those norm-sample categories and dates differ from the census categories, so the percentages are not a direct replication of the census comparison. Ritchie and Tucker-Drob's meta-analysis in Psychological Science of 142 effect sizes and more than 600,000 people put each additional year of education at about one to five IQ points, supporting education as one relevant influence on test performance without identifying how much it caused the particular Canadian and US norm-sample difference. The page on average IQ by education shows the gradient within a single country, and the page on IQ and income sets out what the correlations do and do not support.

Two cautions apply. PIAAC and the Census describe skills and credentials on their own scales, and neither reports an IQ. And the education explanation of the Canadian Wechsler gap is Pearson's attribution after matching, which Miller and colleagues are careful to call plausible rather than proven, and which Dutton and Lynn note can bias a comparison because matching on education partly matches on ability. What is documented is the association; what is estimated is its cause.

10 The Evidence by Design

Read by design rather than by headline, the Canadian evidence sorts into norm samples that define the scale, a comparison of two norm samples on one instrument, an aggregation of uneven samples, a self-selected online table and representative surveys of skills on scales of their own, and the third column of the table below is the one to keep.

The Canadian case shows why reference frames must be named. The same historical standardization participants can receive different scores under Canadian and American tables. That demonstrates a norm dependence; it does not prove that the current Canadian population is exactly 100 or 104.51. Representative educational and adult skills surveys provide additional evidence on their own scales and with their own target populations. Their results should be interpreted alongside sample design and uncertainty rather than merged into a single national IQ.

11 What the Evidence Supports, Stated Narrowly

The evidence supports several distinct statements about Canadian cognitive scores, with limits determined by each design. First, Canadian Wechsler norms center reference performance near 100; a future edition will provide a new reference frame, not a proof that the current population mean is immutable. Second, the 681 participants with complete Full Scale data in the Canadian WAIS-IV standardization of 688 adults scored a mean of 104.51 with a smaller standard deviation when placed on the American norms, a difference replicated across editions, attributed by the publisher mainly to education and demographic composition, reduced to small effect sizes by matching, and disputed in its clinical implications but not in its existence. Third, Becker's 99.62 combines a psychometric summary from ten heterogeneous samples with school assessments. Five Wechsler rows average about 101 after correction; three old records concerning Indigenous participants have substantial interpretation and provenance limits and cannot represent all Indigenous Canadians. Fourth, the 102.09 of the online table describes 7,685 people who chose to sit one test, and it lies about one point from the British and American participant means in that release. Fifth, the sources reviewed here do not establish a comparable current IQ mean for every province; PISA 2025 puts Alberta at the top in science and reading, Quebec at the top in mathematics, Ontario and British Columbia close to the Canadian mean, and the other six provinces below it in every subject, on a scale that measures schooling and with a caution about participation rates printed on the tables.

What the evidence does not support is the reading of 104.5, or 102.5, or 102.09, as Canada's IQ, because the first is a comparison of norm samples, the second adds an assumption to it and the third is a mean of volunteers. It does not establish an overall current adult IQ decline from PISA trends or an overall current rise from the older school studies. The meta-analyses concern different included studies and periods. This source collection also does not supply a comparable current IQ ranking for every province. And it does not support the inference from any of these figures to a person, which the next section takes up.

The narrow statement is more useful than the round number because it tells a reader where to look. For the population, the norm manuals, PISA and the Survey of Adult Skills are the evidence, and each is on its own scale. For a person, the evidence is a score they sat, on a stated scale, with a stated error. The page on average IQ explains what 100 means for the population that defines it, and the page on what a good IQ is explains why the individual question has a different answer.

12 What This Does Not Say About You

A national mean cannot determine an individual Canadian's position within a cognitive score distribution. Under a normal model centered at 100 with a standard deviation of 15, about 68 percent of scores fall between 85 and 115 and about 95 percent between 70 and 130. Those are model proportions, not a fresh count of Canadians. The Wechsler comparison also shows why a report should identify its norms: the same performance can receive different numerical scores under different reference tables. On an SD 15 normal model, 130 corresponds approximately to the 98th percentile; that calculation does not establish eligibility for a society without an accepted assessment. Mensa Canada's admission information specifies its requirements, which should be checked directly before testing. The Mensa test guide, IQ score chart and percentile calculator address different parts of that interpretation.

The direction of inference matters as much as the arithmetic. A provincial PISA score describes fifteen year olds in a sampling frame, not any adult in the province. A norm sample mean of 104.51 on foreign tables describes the sample, not a reader who was never in it. An online table describes the people who took that test. And a Full Scale score, from any instrument, summarizes a profile that can be uneven, as the page on fluid versus crystallized intelligence explains.

We sell ACIS, a paid online assessment, so this paragraph carries a commercial interest. The Full Scale form contains 20 subtests across six primary domains, with the reference framework and score evidence documented in the technical manual. Use that documentation for the form and version actually taken; a reliability estimate from a different form should not be transferred automatically. ACIS is self-administered, available in English, and is not a clinical diagnostic instrument. It does not promise a signed clinical report or acceptance for any outside purpose on this page. A Canadian user should evaluate the language, reference frame and administration conditions rather than assume equivalence with Pearson's Canadian tables. For a clinical question or formal requirement, confirm the appropriate professional assessment with the receiving body. For personal exploration, read the current product details and prices before purchasing. An individual result can provide evidence about assessed performance; a country summary cannot supply that result merely from nationality.

13 Sources Behind This Page

Every figure above is traceable to one of the following, and each is linked at the point where it is used. Dataset values for Canada and the comparison countries were read from the NIQ dataset file downloaded from View on IQ, and the calculations described as ours were performed on that file and on the published figures cited. Product details and prices were read on Pearson Canada's, Mensa Canada's and the International IQ Test's pages on September 20, 2026 and will change.

• Pearson Clinical Assessment Canada. Product pages for the WAIS-IV CDN, the WAIS-IV CDN-F, the WISC-V CDN and the WAIS-5 US version, and the WAIS-5 Canadian edition standardization study page, read September 20, 2026.
• Pearson Clinical Assessment Canada. Differences in Canadian and U.S. norms using the WAIS-IV, technical note dated December 2014. pearsonclinical.ca, read September 20, 2026.
• Miller J L, Weiss L G, Beal A L, Saklofske D H, Zhu J and Holdnack J A. Intelligent use of intelligence tests: Empirical and clinical support for Canadian WAIS-IV norms. Journal of Psychoeducational Assessment, 2015, volume 33, issue 4, pages 312 to 328.
• Harrison A G, Armstrong I T, Harrison L E, Lange R T and Iverson G L. Comparing Canadian and American normative scores on the Wechsler Adult Intelligence Scale, Fourth Edition. Archives of Clinical Neuropsychology, 2014, volume 29, issue 8, pages 737 to 746; and Harrison A G, Holmes A, Silvestri R and Armstrong I T. Implications for educational classification and psychological diagnoses using the Wechsler Adult Intelligence Scale, Fourth Edition with Canadian versus American norms. Journal of Psychoeducational Assessment, 2015, volume 33, issue 4, pages 299 to 311.
• Dutton E and Lynn R. The Canadian IQ calculated from the standardization of the WAIS IV. Open Differential Psychology, August 20, 2014.
• Becker D. The NIQ dataset, version 1.3.5, released October 25, 2023. viewoniq.org, read September 20, 2026; with the earlier national values from Lynn R and Vanhanen T, IQ and the Wealth of Nations, Praeger, 2002, and the book built on an earlier version of the file, Lynn R and Becker D, The Intelligence of Nations, Ulster Institute for Social Research, 2019.
• The studies behind the Canadian rows of the dataset: Lange R T, Schoenberg M R, Woodward T S and Brickell T A, Development of the WAIS-III estimate of premorbid ability for Canadians (EPAC), Archives of Clinical Neuropsychology, 2005, volume 20, issue 8, pages 1009 to 1024; Georgas J, van de Vijver F J R, Weiss L G and Saklofske D H, A cross-cultural analysis of the WISC-III, in Culture and Children's Intelligence, Academic Press, 2003, pages 277 to 313; Bouchard M F and colleagues, Intellectual impairment in school-age children exposed to manganese from drinking water, Environmental Health Perspectives, 2011, volume 119, issue 1, pages 138 to 143; Boucher O and colleagues, The relation of lead neurotoxicity to the event-related potential P3b component in Inuit children from arctic Quebec, NeuroToxicology, 2009, volume 30, issue 6, pages 1070 to 1077; Flores M B and Evans G T, Some differences in cognitive abilities between selected Canadian and Filipino students, Multivariate Behavioral Research, 1972, volume 7, issue 2, pages 175 to 191; John J St, Krichev A and Bauman E, Northwestern Ontario Indian children and the WISC, Psychology in the Schools, 1976, volume 13, issue 4, pages 407 to 411.
• International IQ Test. Average IQ by country, table dated January 1, 2026, as used by the ACIS country ranking page. international-iq-test.com, read September 20, 2026.
• Flynn J R. Massive IQ gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 1987, volume 101, issue 2, pages 171 to 191.
• Trahan L H, Stuebing K K, Fletcher J M and Hiscock M. The Flynn effect: A meta-analysis. Psychological Bulletin, 2014, volume 140, issue 5, pages 1332 to 1360.
• Pietschnig J and Voracek M. One century of global IQ gains: A formal meta-analysis of the Flynn effect (1909 to 2013). Perspectives on Psychological Science, 2015, volume 10, issue 3, pages 282 to 306.
• Rindermann H. The g-factor of international cognitive ability comparisons: the homogeneity of results in PISA, TIMSS, PIRLS and IQ tests across nations. European Journal of Personality, 2007, volume 21, issue 5, pages 667 to 706.
• Council of Ministers of Education, Canada. Measuring Up: Canadian Results of the OECD PISA 2025 Study, the performance of Canadian 15 year olds in science, reading and mathematics. cmec.ca, 2026; and Measuring Up: Canadian Results of the OECD PISA 2022 Study. cmec.ca, 2023.
• Statistics Canada. Literacy, numeracy and adaptive problem solving skills of Canadians: results from the 2022 Programme for the International Assessment of Adult Competencies, The Daily, December 10, 2024; and Canada leads the G7 for the most educated workforce, The Daily, November 30, 2022, with table 98-10-0384-01, Highest level of education by census year.
• Ritchie S J and Tucker-Drob E M. How much does education improve intelligence? A meta-analysis. Psychological Science, 2018, volume 29, issue 8, pages 1358 to 1369.
• Mensa Canada. Take the test, admission testing page. mensa.ca, read September 20, 2026.

The Standards for Educational and Psychological Testing (AERA, APA and NCME, 2014) and the APA Ethical Principles of Psychologists and Code of Conduct, Assessment standards provide the interpretive framework: evidence must support the use, the reference group and limitations must be stated, and assessment results require appropriate explanation. Citing these standards does not imply their authors endorse ACIS or another product.

14 Frequently Asked Questions