Human handedness: A meta-analysis

TL;DR: It is argued that the same evolutionary mechanisms should apply across geographical regions to maintain the roughly 1:10 ratio, while cultural factors, such as pressure against left-hand use, moderate the magnitude of the prevalence of left-handedness.

Abstract: Across time and place, right hand preference has been the norm, but what is the precise prevalence of left- and right-handedness? Frequency of left-handedness has shaped and underpinned different fields of research, from cognitive neuroscience to human evolution, but reliable distributional estimates are still lacking. While hundreds of empirical studies have assessed handedness, a large-scale, comprehensive review of the prevalence of handedness and the factors that moderate it, is currently missing. Here, we report 5 meta-analyses on hand preference for different manual tasks and show that left-handedness prevalence lies between 9.3% (using the most stringent criterion of left-handedness) to 18.1% (using the most lenient criterion of nonright-handedness), with the best overall estimate being 10.6% (10.4% when excluding studies assessing elite athletes' handedness). Handedness variability depends on (a) study characteristics, namely year of publication and ways to measure and classify handedness, and (b) participant characteristics, namely sex and ancestry. Our analysis identifies the role of moderators that require taking into account in future studies on handedness and hemispheric asymmetries. We argue that the same evolutionary mechanisms should apply across geographical regions to maintain the roughly 1:10 ratio, while cultural factors, such as pressure against left-hand use, moderate the magnitude of the prevalence of left-handedness. Although handedness appears as a straightforward trait, there is no universal agreement on how to assess it. Therefore, we urge researchers to fully report study and participant characteristics as well as the detailed procedure by which handedness was assessed and make raw data publicly available. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Summary

Introduction

• Frequency of left-handedness has shaped and underpinned different fields of research, from cognitive neuroscience to human evolution, but reliable distributional estimates are still lacking.
• While hundreds of empirical studies have assessed handedness, a large-scale, comprehensive review of the prevalence of handedness and the factors which moderate it, is currently missing.

Public significance statements

• ● To date, this meta-analysis is the largest reported study to estimate the prevalence of left hand preference for different manual tasks across geographical areas (n = 2,396,170 individuals).
• More data is needed for individuals with less represented ancestries.
• When three handedness categories are given (left-handed, mixed-handed, right- handed), the best estimate for the prevalence of mixed-handedness is 9.33%, a number almost as large as the prevalence of left-handedness.
• Hand preference measurement moderates the estimated prevalences of left- and right- handedness.
• Moreover, studies need to fully report study characteristics, such as instrument used to measure handedness (including questionnaire length and individual item content), response format, classification scheme, country in which the study took place, as well as population characteristics, such as sex, age, ancestry, educational and sporting level of the participants, ideally by uploading raw data in open-access repositories.

Human handedness: A meta-analysis

• The population-level preferential use of the right hand has been the case at least since the days of Homo habilis, the precursor of modern Homo sapiens, two million years ago (Frayer et al., 2016; McManus, 2002).
• It has been suggested that one important methodological tool to avoid this problem and to identify true effects in psychological research are large-scale meta-analyses (Maxwell, Lau, & Howard, 2015).
• A large-scale, comprehensive review of the prevalence of handedness and its moderators is currently lacking, despite the wealth of studies investigating handedness.
• Small study bias may be caused by publication bias (also termed ascertainment bias), but could also be caused by other factors such as systematic differences in study quality between small and large studies.

Why handedness matters

• Understanding handedness and the prevalence of its different manifestations (i.e., left- , non-right-, mixed-, and right-handedness, as well as ambidexterity) can be very informative in a range of research fields, such as neuroscience, genetics, psychiatry, cognitive psychology, psychoneuroendocrinology, as well as evolutionary biology.
• Factors that potentially moderate handedness prevalence Suggesting a genetic basis for handedness does not eliminate possible gene- environment interactions.
• But even amongst the different hand preference inventories one can find differences in patterns of distributions of the prevalence of right- and lefthandedness (Holder, 1992).
• Unfortunately, these cut-off schemes often vary widely between different studies.

Scope of the present study

• Here the authors report five large-scale meta-analyses, which integrate research findings from the broad field of handedness, in order to produce a reliable estimate of the prevalence of handedness categories, namely left-, non-right-, and mixed-handedness versus righthandedness.
• The authors further investigate the effect of factors that have been suggested to moderate the prevalence of handedness.
• Finally, the heterogeneity amongst the included studies and the presence of small study bias were assessed.

Study Selection

• The authors search strategy was based on the strategy followed by Papadatou-Pastou, Martin, Munafò, and Jones (2008).
• All their included studies were entered to represent the years 1927- 2007, but was expanded to include steps d and e (described above).
• Data collection ended in June 2018 and 200 studies were included in the meta-analysis.
• Details about the method of literature search and data extraction between October 2007 and June 2018 are shown in Figure 1.
• The PRISMA statement (Moher, Liberati, Tetzlaff, Altman, & The PRISMA Group, 2009) on reporting items for systematic reviews and meta-analyses was followed.

Inclusion and exclusion criteria

• The following criteria were set for inclusion of a dataset in the systematic review: (a) Participants: Unless stated otherwise, the authors assumed that participants were healthy, heterosexual singletons.
• Data from participants acting as controls to twins or pathological populations were used (e.g., twin controls: Heikkilä et al.
• Different studies classified their participants into different handedness groups.
• The most used classifications were Right-Mixed-Left (R-M-L), Right-Left (R-L), and RightnonRight (R-nonR).
• When two or more samples from different geographical areas (e.g., de la Fuente, Casasanto, Román, & Santiago, 2015) or from different age groups (Ocklenburg et al., 2016) were reported in the same article, they were treated as separate datasets.

Moderator variables

• The variables that were extracted to test for possible moderating effects on the prevalence of handedness were the following (for a more detailed description see PapadatouPastou et al., 2008): Instrument.
• The following schemes were used: Right-Mixed-Left (R-M-L), Right-Left (R-L), and Right-nonRight (R-nonR).
• Rarely was information about ancestry reported, but it was rather inferred from the location in which the testing took place, resulting in three groupings: sub-Saharan African, European, and East Asian.
• In order to test for the possible moderating effects of the year of publication of the studies, mean age of participants, and the length of the questionnaire (i.e., number of questionnaire items used) numerical values were used.

Statistical analysis

• Meta-analysis was carried out in R using the robumeta package (Fisher & Tipton, 2015).
• The variation in the classification schemes used in the original studies did not allow for a single, overall analysis to take place without losing important information.
• The groups, which were analyzed in separate meta-analyses, were as follows: 1. Left-handedness : 2. Non-right-handedness: Non-right-handers correspond to participants who were classified as non-right-handers in datasets where an R-nonR classification was employed.
• In order to investigate the presence of small study bias the authors used the funnel plot graphical test (funnel function), Egger’s regression test (regtest function), and Duval and Tweedie’s (2000) trim and fill method (trimfill function) of the R metafor package (Viechtbauer, 2010), also known as Step 5.

Moderating variables analysis

• Because of the heterogeneity detected among studies, the moderating effects of the previously described variables were tested within the left-handedness comparison, which was the most inclusive.
• The authors suggest that study characteristics, such as instrument used to measure handedness (including questionnaire length and individual item content), response format, classification scheme, country in which the study took place, as well as population characteristics, such as sex, age, ancestry, and educational level of the participants, are reported in all handedness studies.

Figure captions

• Flow diagram for the search (October 2007 – June 2018) and inclusion criteria for studies in the systematic review and meta-analysis.
• The figure was created according to the guidelines of the PRISMA statement (Moher et al., 2009).
• Data were required to have been broken down by sex in a comprehensive way.
• Handedness was required to have been measured in terms of preference, not performance.

Figures

Table 3. Overall Incidence of Left-Handedness in the Data Sets

Table 4. Moderator Variables Extracted From Each Study

Table 2. Incidence of Left-Handedness in the Different Levels of the Moderator Variables Within the Left-Handedness (Total) Comparison

Figure 5. Forest plot for the left-handedness (total) comparison grouped by instruments used to assess handedness. In the plot the 95% confidence interval for each study is represented by a horizontal line and the point estimate is represented by a dot. The dashed line indicates the overall estimate of left-handedness prevalence.

Figure 4. Forest plot for the left-handedness (total) comparison grouped by classification. In the plot the 95% confidence interval for each study is represented by a horizontal line and the point estimate is represented by a dot. The dashed line indicates the overall estimate of lefthandedness prevalence.

Figure 6. Scattergram visualizing the event rates of individual studies against their respective standard errors.

Figure 3. Forest plot for the left-handedness (total) comparison grouped by year of publication (categorical). In the plot the 95% confidence interval for each study is represented by a horizontal line and the point estimate is represented by a dot. The dashed line indicates the overall estimate of left-handedness prevalence.

Table 1. Details of Studies Included in the Meta-Analyses

Figure 7. Forest plot for the left-handedness (total) comparison grouped by geographical ancestry. In the plot the 95% confidence interval for each study is represented by a horizontal line and the point estimate is represented by a dot. The dashed line indicates the overall estimate of left-handedness prevalence.

Full text

HUMAN HANDEDNESS: A META-ANALYSIS
1
Running head: HUMAN HANDEDNESS: A META-ANALYSIS
Human handedness: A meta-analysis
Marietta Papadatou-Pastou
1
, Eleni Ntolka
1
, Judith Schmitz
2
, Maryanne Martin
3
,
Marcus R. Munafò
4,5
, Sebastian Ocklenburg
6
*, Silvia Paracchini
2
*
1
School of Education, National and Kapodistrian University of Athens, Greece
2
School of Medicine, University of St Andrews, St. Andrews, UK
3
Department of Experimental Psychology, University of Oxford, UK
4
MRC Integrative Epidemiology Unit at the University of Bristol, UK
5
School of Psychological Science, University of Bristol, UK
6
Biopsychology, Department of Psychology, Institute of Cognitive Neuroscience,
Ruhr University Bochum, Bochum, Germany
Address for correspondence: Dr. Marietta Papadatou-Pastou, School of Education, National
and Kapodistrian University of Athens, 13A Navarinou Str., Athens 160 80, Greece.
Email: marietta.papadatou-pastou@seh.oxon.org, Tel: +30 210 3641712
*Both authors contributed equally
© 2020, American Psychological Association. This paper is not the copy of
record and may not exactly replicate the final, authoritative version of the
article. Please do not copy or cite without authors' permission. The final article
is available via its DOI: 10.1037/bul0000229

HUMAN HANDEDNESS: A META-ANALYSIS
2
Abstract
Across time and place, right hand preference has been the norm, but what is the precise
prevalence of left- and right-handedness? Frequency of left-handedness has shaped and
underpinned different fields of research, from cognitive neuroscience to human evolution, but
reliable distributional estimates are still lacking. While hundreds of empirical studies have
assessed handedness, a large-scale, comprehensive review of the prevalence of handedness
and the factors which moderate it, is currently missing. Here, we report five meta-analyses
on hand preference for different manual tasks and show that left-handedness prevalence lies
between 9.3% (using the most stringent criterion of left-handedness) to 18.1% (using the
most lenient criterion of non-right-handedness), with the best overall estimate being 10.6%
(10.4% when excluding studies assessing elite athletes’ handedness). Handedness variability
depends on (a) study characteristics, namely year of publication and ways to measure and
classify handedness, and (b) participant characteristics, namely sex and ancestry. Our
analysis identifies the role of moderators which require taking into account in future studies
on handedness and hemispheric asymmetries. We argue that the same evolutionary
mechanisms should apply across geographical regions to maintain the roughly 1:10 ratio,
while cultural factors, such as pressure against left-hand use, moderate the magnitude of the
prevalence of left-handedness. Although handedness appears as a straightforward trait, there
is no universal agreement on how to assess it. Therefore, we urge researchers to fully report
study and participant characteristics as well as the detailed procedure by which handedness
was assessed and make raw data publicly available.
Keywords: Handedness; Meta-Analysis; Laterality; Hand Preference; Cerebral Asymmetries

HUMAN HANDEDNESS: A META-ANALYSIS
3
Public significance statements
● To date, this meta-analysis is the largest reported study to estimate the prevalence of
left hand preference for different manual tasks across geographical areas (n =
2,396,170 individuals). It shows that the best estimate for the prevalence of left-
handedness is 10.6%. However, this value varies between 9.3% and 18.1%,
depending on how handedness is measured.
● The same evolutionary mechanisms should apply to participants of different
geographical ancestries to maintain the roughly 1:10 ratio of left- versus right-handers
found worldwide. The exact prevalence of left hand preference is moderated by
cultural factors, primarily pressure to change writing hand, possibly due to direct
instructions by parents and teachers and also through non-explicit model learning.
More data is needed for individuals with less represented ancestries.
● When three handedness categories are given (left-handed, mixed-handed, right-
handed), the best estimate for the prevalence of mixed-handedness is 9.33%, a
number almost as large as the prevalence of left-handedness. This highlights the
importance of taking this group into account in future handedness studies.
● Hand preference measurement moderates the estimated prevalences of left- and right-
handedness. We urge researchers to define universal criteria for measuring hand
preference (short questionnaires, reporting both writing hand and Edinburgh
Handedness Inventory [EHI] scores, and reporting at least two classifications, e.g. R-
L and R-M-L), as measurement imprecision and/or heterogeneity affects the estimated
prevalence. Moreover, studies need to fully report study characteristics, such as
instrument used to measure handedness (including questionnaire length and individual
item content), response format, classification scheme, country in which the study took

HUMAN HANDEDNESS: A META-ANALYSIS
4
place, as well as population characteristics, such as sex, age, ancestry, educational and
sporting level of the participants, ideally by uploading raw data in open-access
repositories. Detailed reporting is essential to compare effectively different studies as
well as to encourage good study design.

HUMAN HANDEDNESS: A META-ANALYSIS
5
Human handedness: A meta-analysis
Laterality is a general principle of functional organization in vertebrates (e.g.,
Bisazza, Rogers, & Vallortigara, 1998; Güntürkün & Ocklenburg, 2017; Ocklenburg, Isparta,
Peterburs, & Papadatou-Pastou, 2019; Rogers, 2008; Vallortigara & Rogers, 2005). Homo
sapiens have evolved laterality in a unique form within our primate lineage (Uomini & Ruck,
2018). The population-level preferential use of the right hand has been the case at least since
the days of Homo habilis, the precursor of modern Homo sapiens, two million years ago
(Frayer et al., 2016; McManus, 2002). Recent estimations place the emergence of the right-
hand bias in the course of the last seven million years (Uomini & Ruck, 2018).
Despite the fact that population-level right-hand preference is well established, the
precise magnitude of the percentages of right- and left-handedness still remains to be
elucidated. Handedness prevalence is indeed a point of dispute among different studies. Some
of these differences might be explained by small sample sizes in individual studies, a problem
that has been identified as one of the reasons for the current replication crisis in psychology.
In addition to small sample sizes, inconsistent results are likely to be driven by publication
bias, p-hacking, and heterogeneity in how handedness is measured. It has been suggested that
one important methodological tool to avoid this problem and to identify true effects in
psychological research are large-scale meta-analyses (Maxwell, Lau, & Howard, 2015). A
large-scale, comprehensive review of the prevalence of handedness and its moderators is
currently lacking, despite the wealth of studies investigating handedness.
Solving the riddle of the prevalence of handedness is not an easy task, because of the
large number of published studies on handedness. For example, entering the key word
'handedness' in PubMed in December 2019 resulted in 60,868 hits. A meta-analysis lends

Frequently asked questions

Q1. What are the contributions in this paper?

In this paper, the authors presented the results of a study at the National and Kapodistrian University of Athens, Greece and the University of St Andrews, St. Andrews. 

Q2. How many studies will need to be imputed to the left of the mean?

According to the trim and fill test, seven studies (SE = 3.37) will need to be imputed to the left of the mean, corresponding to lower non-right-handedness rates, in order for the funnel plot to be symmetrical. 

Q3. What was the need for the data to be broken down by handedness groups?

Data needed to be broken down by handedness groups and sex in a format that provided arithmetic data that could be used in the present analysis (i.e., not reporting only laterality quotients or p-values or providing only graphical representations of data). 

Q4. What is the reason for the higher prevalence of left-handedness in sporting elites?

The trend towards higher left-handedness prevalence in sporting elites is possibly due to the nature of interactive sports giving left-handers an advantage as righthanders have less practice with left-handed opponents (Loffing, 2017; Loffing & Hagemann, 2012, 2016; Wood & Aggleton, 1989). 

Q5. What could be the moderating effects of ancestry?

The moderating effects of ancestry could be either due to (a) genetics, (b) prenatal testosterone levels and its geographical variations (Raymond & Pointier, 2004), or (c) cultural factors. 

Q6. Why was the writing hand included in the meta-analysis?

Due to lack of usable arithmetic data, out of the 10 items of the EHI only the writing hand item was taken into account for the meta-analysis. 

Q7. What is the common criterion for assignment?

In the first case, the categories are usually 'right' and 'left', with writing hand being the most common criterion for group assignment. 

Q8. What is the influential theory of the evolution of left-handedness?

Another influential theory is that of the gradual easing of cultural pressures against sinistrality (Schachter, Ransil, & Geschwind, 1987). 

Q9. What is the estimate for the prevalence of left-handedness?

When three handedness categories are given (left-handed, mixed-handed, right-handed), the best estimate for the prevalence of mixed-handedness is 9.33%, a number almost as large as the prevalence of left-handedness.