ArticlePDF Available

Edinburgh Handedness Inventory - Short Form: A revised version based on confirmatory factor analysis

Authors:

Abstract and Figures

While the Edinburgh Handedness Inventory has been widely used, there have been few studies assessing its factorial validity. There is evidence that the original instructions and response options are difficult to understand. Using simplified instructions and response options, the Edinburgh Handedness Inventory was administered on a sample of 1514 participants using an online questionnaire. In accordance with previous research, a model of the 10-item inventory had poor fit for the data. This study also detected model misspecification in the previously-proposed 7-item modification. A 4-item Edinburgh Handedness Inventory - Short Form had good model fit with items modelled as both continuous and ordinal. Despite its brevity, it showed very good reliability, factor score determinacy, and correlation with scores on the 10-item inventory. By eliminating items that were modelled with considerable measurement error, the short form alleviates the concern of the 10-item inventory over-categorising mixed handers. Evidence was found for factorial invariance across level of education, age groups, and regions (USA and Australia/New Zealand). There generally appeared to be invariance across genders for the 4-item inventory. The proposed Edinburgh Handedness Inventory - Short Form measures a single handedness factor with an inventory that has brief and simple instructions and a small number of items.
Content may be subject to copyright.
_____________________________
Correspondence: Department of Psychology, University of British Columbia, 2136 West
Mall, Vancouver, BC, V6T 1Z4, Canada; email: jaimie@jaimieveale.com
This is an Accepted Manuscript of an article published in Laterality: Asymmetries of Body, Brain and Cognition 2014 19(2),
164-177, available online: http://www.tandfonline.com/10.1080/1357650X.2013.783045
EDINBURGH HANDEDNESS INVENTORYSHORT FORM: A
REVISED VERSION BASED ON CONFIRMATORY FACTOR
ANALYSIS
Jaimie F. Veale
Department of Psychology, University of British Columbia.
ABSTRACT
While the Edinburgh Handedness Inventory has been widely used, there have been few
studies assessing its factorial validity. There is evidence that the original instructions and
response options are difficult to understand. Using simplified instructions and response
options, the Edinburgh Handedness Inventory was administered on a sample of 1514
participants using an online questionnaire. In accordance with previous research, a model of
the 10-item inventory had poor fit for the data. This study also detected model
misspecification in the previously-proposed 7-item modification. A 4-item Edinburgh
Handedness Inventoryshort form had good model fit with items modeled as both
continuous and ordinal. Despite its brevity, it showed very good reliability, factor score
determinacy, and correlation with scores on the 10-item inventory. By eliminating items that
were modeled with considerable measurement error, the short form alleviates the concern of
the 10-item inventory over-categorizing mixed handers. Evidence was found for factorial
invariance across level of education, age groups, and regions (USA and Australia/New
Zealand). There generally appeared to be invariance across genders for the 4-item inventory.
The proposed Edinburgh Handedness Inventoryshort form measures a single handedness
factor with an inventory that has brief and simple instructions and a small number of items.
KEY WORDS: handedness, Edinburgh Handedness Inventory, confirmatory factor analysis,
invariance testing, psychometric properties
Jaimie Veale
2
Handedness, or the preference to use one hand more than the other, is usually
measured by questionnaire in the social sciences. The Edinburgh Handedness Inventory
(Oldfield, 1971) is the most widely used of these questionnaires (Fazio, Coenen, & Denney,
2012). Oldfield’s 1971 article currently has almost 11000 citations in Scopus. To complete
the Edinburgh Handedness Inventory, respondents endorse hand preference for ten everyday
tasks. In the original inventory, participants are given a list of the tasks with adjacent “left”
or “right” columns. They are asked write “+” in the appropriate corresponding column. If the
preference is so strong that they “would never try to use the other hand unless absolutely
forced to” they are to write “++” instead. If they are “really indifferent” they are instructed to
put an “+” in both columns (Oldfield, 1971, p. 112).
Despite its widespread usage, there have been few studies assessing the factorial
validity of the Edinburgh Handedness Inventory. Such assessment is important because it
tests that the inventory measures the unidimensional handedness factor it purports to. Studies
that have been conducted have generally noted that some of the questionnaire items are
problematic (Bryden, 1977; Dragovic, 2004; McFarland & Anderson, 1980; S. M. Williams,
1986). From exploratory factor analysis, Bryden found three items (opening box, using
broom, and striking a match) loaded onto a single factor “in which there was considerable
disagreement among right-handers as to which hand was generally used, and items in which
the subject had to think very carefully before giving an answer” (p. 621). Also from
exploratory factor analysis, McFarland and Anderson noted “poor” loading of the knife,
broom, and box lid items on a single handedness factor. They reported evidence for factor
loading stability across age and gender, and test-retest for most of the items, but noted
instability on the box, broom, match, and scissors item. However, McFarland and Anderson
did not test whether allowing these factor loadings to differ between groups significantly
improved the model fit and they did not test the equality of item intercepts or error variances
which are requirements for testing whether findings of group mean differences can be
accounted for by measurement biases (Gregorich, 2006).
Confirmatory factor analysis is a powerful tool for assessing the factorial validity of a
questionnaire. Advantages of confirmatory factor analysis are that it allows modeling of
error variance, testing for item uniqueness, and testing for acceptable fit of the factor
structure (Brown, 2006). Two studies have tested the Edinburgh Handedness Inventory using
confirmatory factor analysis. From a community sample of 203, Dragovic (2004) found it
was redundant to include both the writing and drawing items, as these appeared to be almost
perfectly collinear, and that the broom and box-lid were modeled with a high proportion of
Edinburgh Handedness InventoryShort Form
3
error variance. He concluded that this high error variance was due to these items measuring
another factor or ambiguity of item interpretation. Dragovic proposed a 7-item revised
inventory with these three items removed. Milenkovic and Dragovic (2012) replicated these
findings on a sample of 1224 high school students.
There were limitations of these previous two confirmatory factor analysis studies.
These studies used the maximum likelihood estimation technique that assumes indicators are
continuous. However, with 5-point Likert response options, an estimation technique for
ordered-categorical data is also appropriate (Kline, 2011). These two previous studies
administered the Edinburgh Handedness Inventory using the original instructions. These
instructions have been criticized as “somewhat lengthy and confusing” (Fazio et al., 2012, p.
71) and there is evidence that the majority of male inmates who were administered the
inventory did not understand or follow them correctly (Fazio et al., 2012). Fazio et al.
suggested a Likert scale adaptation of the Edinburgh Handedness Inventory could markedly
improve instruction adherence.
METHOD
Participants
Participants were recruited for an internet-based study examining the development of
gender identity and sexuality (Veale, Clarke, & Lomax, 2010) through lesbian, gay, bisexual,
and transgender related online forums and mailing lists, Google online advertising, and a
press release. The Edinburgh Handedness Inventory was completed by 1514 participants.
Participants’ demographics are outlined in Table 1. Age ranged from 16 to 81. A
significant proportion of participants had gender identities not consistent with their biological
sex (e.g. transsexual, transgender). For invariance testing between genders, 160 biological
males and 266 biological females who did not report having gender-variant identities were
used.
Measure
The Edinburgh Handedness Inventory (Oldfield, 1971) was administered with revised
instructions and response options. The instructions given were “Please indicate your
preferences in the use of hands in the following activities. Some of the activities require both
hands. In these cases the part of the task, or object, for which hand preference is wanted is
Jaimie Veale
4
indicated in brackets”. Response options given were “always right”, “usually right”, “both
equally”, “usually left”, and “always left”.
Table 1. Participants’ demographics
Ethnicity
%
Residence
%
Level of education
%
White/Caucasian
92
USA
60
3 years high school
7
East Asian
3
New Zealand
18
4 years high school
10
Hispanic/Latino
3
Great Britain
8
5 years high school
11
Indigenous American
3
Canada
5
Diploma/trade qualification
20
Black/African
2
Australia
2
Bachelor’s degree
31
South/other Asian
2
Other
7
Master’s degree
14
Other
2
Doctoral degree
5
Data Analysis
Confirmatory factor analysis of the Edinburgh Handedness Inventory was conducted
using Mplus software version 5.1 (Muthén & Muthén, 2008). Some authors have suggested
5-point response scales be treated as ordinal (Kline, 2011), while others have suggested
treatments as continuous (Blunch, 2008). Analyses for this article were conducted using both
Yuan-Bentler robust maximum likelihood estimation and mean- and variance-adjusted
weighted least squares (WLSMV) estimation. The former estimates and fit indices are
adjusted to allow for missing data and variations of multivariate normality. The latter allows
for ordinal categories, missing data, and uses a polychoric correlation matrix (Kline, 2011).
Model fit was assessed using absolute fit indices: Satorra-Bentler χ² likelihood ratio
(SBχ²), RMSEA, and SRMR and relative fit indices: CFI and TLI. Model misspecification
was detected by a statistically significant p value less than .05 on the χ² test. CFI and TLI
values lower than .9, RMSEA values greater than .05 and SRMR values greater than .08 were
also used as indicators that the proposed model did not fit the data (Kline, 2011).
Reliability was estimated using Cronbach’s (1951) α and Raykov’s (1997) factor ρ
which is a composite reliability coefficient that is calculated as the ratio of variance explained
by the factor to the total variance. Factor score determinacy was also estimated by
calculating the squared multiple correlation of the proposed indicators for predicting the
handedness factor. This gives information about the extent to which the true factor score can
be determined in the model (Grice, 2001). This is useful because while confirmatory factor
Edinburgh Handedness InventoryShort Form
5
analysis models with a small number of indicators may be more likely to fit the data, they are
also more likely to have factor indeterminacy (Brown, 2006).
In accordance with Milenkovic and Dragovic’s (2012) study, laterality quotients were
calculated. These ranged from -100 (left handed) to 100 (right handed). Participants with a
laterality quotient score between -60 and 60 (inclusive) were classified as mixed handers.
Testing for parallel indicators was also conducted to test whether each item measures the
handedness construct equally. Such testing is desirable to understand whether summing
scores on each item equally (as per the laterality quotient) will accurately estimate a person’s
handedness.
The inventory’s stability across a number of groups was also assessed using factorial
invariance testing. The majority of participants sampled in this thesis lived in two regions:
the USA (60%) or Australia and New Zealand (20%). Invariance testing was conducted
between these groups to test for differences in item meaning for participants living in these
regions. The median age of participants was 37. Invariance testing was conducted between
those above and below the median age. Finally, invariance testing was conducted between
males and females (participants with gender identities atypical to their biological sex are
removed from this analysis).
Metric invariance and scalar invariance were tested by comparing these to a model
where these invariance constraints were not imposed (the unconstrained model). A
statistically significant change in scaled difference χ² likelihood ratio test may indicate scale
noninvariance (Kline, 2011). However, because this test can be over-sensitive when sample
size is large, Cheung and Rensvold’s (2002) criterion of a decrease in CFI of greater than .01
is evidence for measurement noninvariance was also used.
Data were missing for 1% of responses. Analyses were conducted using Mplus’s full
information maximum likelihood method (Asparouhov & Muthén, 2010).
RESULTS
Model Testing and Selection
As outlined in Tables 2 and 3, confirmatory factor analysis using all ten items of the
Edinburgh Handedness Inventory elicited similar concerns to previous studies. The model
had poor performance on all fit indices with the exception of the CFI and TLI using WLSMV
estimation. There was collinearity between the drawing and writing items, r = .97, and high
proportion of residual error for the broom (.64 using robust maximum likelihood, .44 using
Jaimie Veale
6
WLSMV estimation), knife (.54, .35), and lid opening (.54, .34) items. The covariance
matrix for the 10 items is given in supplementary electronic data. Dragovic’s proposed 7-
item inventory, in which the drawing, broom, and box lid opening items are removed, had
signs of model misspecification on the χ² likelihood ratio using both estimation techniques
and also the RMSEA when using WLSMV estimation, so a revised model was also tested.
Table 2. Fit statistics, and reliability and factor score determinacy estimates for versions of
the Edinburgh Handedness Inventory with robust maximum likelihood estimation
Model
df
p
CFI
TLI
RMSEA
SRMR
ρ
α
FSD
10 items
35
< .001
.84
.79
.145
.05
.95
.95
.98
7 items
14
< .001
.99
.98
.047
.02
.95
.95
.98
4 items
2
.819
1.00
1.00
.000
.00
.93
.93
.97
4 items, parallel
indicators
5
< .001
.98
.97
.075
.08
.93
.97
Note. N = 1514; ρ = Raykov’s composite reliability; FSD = factor score determinacy
Table 3. Fit statistics, and reliability and factor score determinacy estimates for versions of
the Edinburgh Handedness Inventory with WLSMV estimation
Model
χ²
df
p
CFI
TLI
RMSEA
10 items
271.87
21
< .001
.99
1.00
.089
7 items
103.67
14
< .001
1.00
1.00
.065
4 items
1.97
2
.373
1.00
1.00
.000
4 items, parallel indicators
158.53
4
< .001
.99
1.00
.160
Note. N = 1514
In selecting a revised model, the following were considered:
- As outlined above, previous research has noted problems with the knife, striking match,
and scissors items;
- although this past evidence is more limited for the knife and scissors items, the knife item
was modeled with significant error variance in the previous analysis and scissor use
preference may be effected by practice with the tools available (i.e. many modern scissors
are designed for right-handed use); and
- the match and knife items are the only remaining items pertaining to tasks which require
two handsremoving these will simplify the scale.
Edinburgh Handedness InventoryShort Form
7
Given these considerations, the knife, striking match, and scissors items were
removed. Tables 2 and 3 show the resultant 4-item model had adequate fit for the on all fit
tests using both estimation techniques. Constraining factor loadings to be equal amongst
indicators (parallel indicators) resulted in a model with poor fit using both estimation
techniques. Reliability and factor score determinacy estimates are also given in Table 2.
Parameter estimates for the 4-item model are given in Table 4.
Table 4. Model estimates for the 4-item Edinburgh Handedness Inventory
Parameter
Unstandardized
SE
Standardized
Unstandardized
SE
Standardized
MLR factor loadings
WLSMV factor loadings
Writing
1.00a
__
.89
1.00a
__
.97
Throwing
0.83
.03
.82
0.89
.01
.87
Toothbrush
0.94
.02
.88
0.95
.01
.92
Spoon
1.00
.02
.91
0.97
.01
.94
MLR factor variance
WLSMV factor variance
Handedness
32.29
2.23
1.00
0.95
.01
1.00
Note. SE = standard error. MLR = robust maximum likelihood. a This parameter is fixed
because this item is used as a marker variable, therefore it is not tested for statistical
significance. All other estimates were statistically significant p < .001.
Invariance Testing
Tables 5 and 6 outline the results of invariance testing results. These analyses showed
no significant change towards poorer model fit in the χ² likelihood ratio test. While the
change in this test for between-countries metric invariance achieved statistical significance
(χ²(3) = 8.13, p = .04) when using WLSMV estimation, this represented an improvement in
model fit. Although not detected by the χ² likelihood ratio test, change in CFI for between-
genders invariance testing was notable. This can be explained by constraining the toothbrush
item between genders resulting in a significant scaled difference χ²(1) = 4.08, p = .04,
suggesting toothbrush hand preference has a slightly higher factor loading in males.
Unfortunately, WLSMV between-gender invariance testing could not be conducted because
of a not positive definite correlation matrix, likely due to a low number of “both equally” and
“usually left” response options to the writing variable.
Jaimie Veale
8
Table 5. Invariance testing fit statistics for the 4-item Edinburgh Handedness Inventory using
robust maximum likelihood estimation
Model
SBχ²
df
p
χ²SD
∆df
p
CFI
CFI
Country: USA/Australia or New Zealand
n = 1214
Unconstrained
2.14
4
.710
-
-
-
1.000
-
Metric invariance
2.82
7
.901
0.30
3
.960
1.000
.000
Scalar invariance
10.94
10
.362
4.88
6
.560
.999
.001
Age median split
N = 1514
Unconstrained
4.14
4
.388
-
-
-
1.000
-
Metric invariance
7.17
7
.412
1.24
3
.744
1.000
.000
Scalar invariance
15.54
10
.114
6.31
6
.389
.997
.003
Gender
n = 426
Unconstrained
6.48
4
.166
-
-
-
.994
-
Metric invariance
16.27
7
.023
4.61
3
.203
.977
.017
Scalar invariance
22.54
10
.013
8.45
6
.207
.968
.026
Level of education
n = 1446
Unconstrained
3.96
6
.683
-
-
-
1.000
-
Metric invariance
11.88
12
.455
3.21
6
.782
1.000
.000
Scalar invariance
19.91
18
.338
6.46
12
.891
.999
.001
Note. SD = scaled difference
Laterality Quotient
Figure 1 plots participants’ laterality quotient scores on the 10-item inventory and the
4-item short form. Participants with any missing data were excluded from this analysis. The
relationship between these two scales was Spearman’s ρ = .90 and r2 = .94.
Results of assignment to handedness categories are given in Table 7. The level
agreement of this categorization between the 10-item and 4-item inventories was kappa = .73.
Edinburgh Handedness InventoryShort Form
9
Table 6. Invariance testing fit statistics for the 4-item Edinburgh Handedness Inventory using
WLSMV estimation
Model
χ²
df
p
χ²Difference
∆df
p
CFI
CFI
Country: USA/Australia or New Zealand
n = 1214
Unconstrained
48.82
9
< .001
-
-
-
.998
-
Metric invariance
45.33
10
< .001
8.13
3
.043
.998
.000
Scalar invariance
40.20
11
< .001
6.60
4
.159
.998
.000
Age median split
N = 1514
Unconstrained
52.69
10
< .001
-
-
-
.998
-
Metric invariance
46.26
11
< .001
6.53
3
.089
.998
.000
Scalar invariance
45.28
12
< .001
8.37
4
.079
.998
.000
Level of education
n = 1446
Unconstrained
35.23
17
.005
-
-
-
.999
-
Metric invariance
31.61
19
.035
4.91
5
.427
.999
.000
Scalar invariance
35.39
21
.026
11.68
9
.232
.999
.000
DISCUSSION
This is the first study to use confirmatory factor analysis to examine the validity the
Edinburgh Handedness Inventory when using simplified instructions and a response scale.
The proposed Edinburgh Handedness Inventoryshort form, including instructions and
response scale, is reproduced in the Appendix. Because the remaining items relate to tasks
that only require one hand, only simplified instructions need to be retained. The simpler
instructions and response options, and lower number of items mean the short form is notably
less burdensome to participants.
Table 7. Assignment of categorical handedness groups in the original and short form
Edinburgh Handedness Inventories
Original version (10-items)
Left n (%)
Mixed n (%)
Right n (%)
Total n
Short form (4-items)
Left
87 (95)
28 (6)
0 (0)
115
Mixed
5 (5)
259 (60)
4 (0)
268
Right
0 (0)
146 (34)
867 (100)
1013
Total
92
433
871
1396
Jaimie Veale
10
Figure 1. Scatter plot of laterality quotients for the Edinburgh Handedness Inventory and its
short form
Note. n = 1396. Jitter function used to illustrate overlapping points.
In accordance with previous research, this study found that the 10-item Edinburgh
Handedness Inventory did not adequately measure a single underlying handedness factor.
Unlike previous studies using confirmatory factor analysis by Dragovic (2004; Milenkovic &
Dragovic, 2012) that suggested a 7-item inventory, this study provided evidence that a 7-item
inventory does not adequately fit the data. Although Dragovic (2004) did not find evidence
for model misspecification for the 7-item inventory, this is likely to be because this study’s
sample size (N = 203) lacked the power detect this. This can be surmised because the two
Edinburgh Handedness InventoryShort Form
11
later studies with larger samples (Milenkovic & Dragovic, 2012 and the present study) found
evidence for this misspecification on the χ² likelihood ratio test.
The 4-item inventory was proposed based on consideration of items’ face validity,
performance of items in this study’s initial analysis, and the findings previous factor analyses
(Bryden, 1977; Dragovic, 2004; McFarland & Anderson, 1980; Milenkovic & Dragovic,
2012). This inventory passed the requirements of all of the fit tests and indices when
indicators were modeled as both continuous and ordinal.
The 4-item Edinburgh Handedness Inventoryshort form performed well on further
validation analyses. Despite it being a short scale, it showed very high reliability on
measures of factorial composite reliability and Cronbach’s alpha (Cicchetti, 1994). The
estimate of the quality of factor scores, factor determinacy, was also high. This is noteworthy
because this is an area which can be of concern when using a low number of indicators for
factor estimation (J. S. Williams, 1978). Also, because the findings of this and another recent
study (Milenkovic & Dragovic, 2012) indicated that the inventory does not have parallel
indicators, researchers may prefer to calculate factor scores to more accurately assess
handedness.
When laterality quotients were calculated, the short form also predicted a large
proportion (94%) of the variance of the 10-item inventory. This is around the same
proportion that Milenkovic and Dragovic (2012) found from the same analysis with the 7-
item inventory. The level of agreement of categorization of participants into left, mixed, and
right handers between original and revised versions in this study (kappa = .73) was lower
than in Milenkovic and Dragovic’s study (kappa = .80), but this finding still suggests
“substantial agreement” (Landis & Koch, 1977). Table 7 illustrates that the main source of
disagreement in this study was participants categorized as mixed handers by the 10-item
inventory were categorized as either left or right handers by the 4-item inventory. The
percentage classified as mixed hander in the 10-item inventory was 31% and this decreased to
19% in the 4-item short form. With recent evidence that the Edinburgh Handedness
Inventory overestimates the proportion of mixed-handers (Büsch, Hagemann, & Bender,
2010; Dragovic, Milenkovic, & Hammond, 2008), it seems that by eliminating items with
notable measurement error, the short from inventory alleviates this concern.
This was the first study to test between-group factorial invariance for the Edinburgh
Handedness Inventory using confirmatory factor analysis. Both metric and scalar invariance
was found between those older than and younger than age 37, those with different levels of
education, and those living in the USA versus Australia and New Zealand. This can be
Jaimie Veale
12
interpreted as evidence for the same factor structure across these groups, further supporting
the validity of the short form. These findings are especially important given recent work
suggesting that a notable proportion of inmates, who are likely to have lower levels of
education on average, did not understand or follow the original Edinburgh Handedness
Inventory correctly. There was some indication that the toothbrush item loaded higher on the
factor in males than females. Because this difference was of marginal significance, only
detected on one of the two fit indices, and there is no obvious theoretical explanation for this
finding, replication of this finding would be prudent before attempting to interpret it.
Nevertheless, researchers interested in the study of gender differences in handedness should
be aware of the possibility of factorial noninvariance for this item.
This was the first confirmatory factor analysis study to validate the Edinburgh
Handedness Inventory using simplified instructions and response options. This was also the
first study to model the response options as ordinal and test factorial invariance statistically.
Results of the analyses suggested the revised 4-item Edinburgh Handedness Inventoryshort
form had very good reliability, factor score determinacy, and model fit. It also appeared to
have enhanced performance in handedness categorization. This short form will be of interest
to researchers wanting to measure a single handedness factor with an inventory that has brief
and simple instructions and a small number of items. This research was conducted using an
online questionnaire; future research could assess the Edinburgh Handedness Inventory
short form using pen-and-paper format, and use biomarkers of handedness to further assess
its validity.
ACKNOWLEDGEMENTS
Contributors: Matt Williams critically reviewed the manuscript; Dave Clarke, Terri
Lomax, and Steve Humphries supervised the larger project; Ellen Stephenson gave helpful
comments regarding face validity of the Edinburgh Handedness Inventory.
This research was supported by a Massey University Doctoral Scholarship and a Post
Doctoral Research Fellowship with the assistance of the Government of Canada/ avec l’appui
du gouvernement du Canada.
Edinburgh Handedness InventoryShort Form
13
REFERENCES
Asparouhov, T., & Muthén, B. O. (2010, September 29). Multiple imputation with Mplus.
Retrieved from http://statmodel.com/download/Imputations7.pdf
Blunch, N. J. (2008). Introduction to structural equation modelling using SPSS and AMOS.
London: Sage.
Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York: Guilford
Press.
Bryden, M. P. (1977). Measuring handedness with questionnaires. Neuropsychologia, 15(4-
5), 617624.
Büsch, D., Hagemann, N., & Bender, N. (2010). The dimensionality of the Edinburgh
Handedness Inventory: An analysis with models of the item response theory.
Laterality, 15, 610628. doi:10.1080/13576500903081806
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing
measurement invariance. Structural Equation Modeling, 9, 233255.
Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for evaluating normed and
standardized assessment instruments in psychology. Psychological Assessment, 6,
284290. doi:10.1037/1040-3590.6.4.284
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16, 297334. doi:10.1007/BF02310555
Dragovic, M. (2004). Towards an improved measure of the Edinburgh Handedness
Inventory: A one-factor congeneric measurement model using confirmatory factor
analysis. Laterality, 9, 411419. doi:10.1080/13576500342000248
Dragovic, M., Milenkovic, S., & Hammond, G. (2008). The distribution of hand preference is
discrete: A taxometric examination. British Journal of Psychology, 99, 445459.
doi:10.1348/000712608X304450
Fazio, R., Coenen, C., & Denney, R. L. (2012). The original instructions for the Edinburgh
Handedness Inventory are misunderstood by a majority of participants. Laterality:
Asymmetries of Body, Brain and Cognition, 17(1), 7077.
doi:10.1080/1357650X.2010.532801
Gregorich, S. E. (2006). Do self-report instruments allow meaningful comparisons across
diverse population groups? Testing measurement invariance using the confirmatory
factor analysis framework. Medical Care, 44(11 Suppl 3), S78S94.
doi:10.1097/01.mlr.0000245454.12228.8f
Jaimie Veale
14
Grice, J. W. (2001). Computing and evaluating factor scores. Psychological Methods, 6(4),
430450. doi:10.1037/1082-989X.6.4.430
Kline, R. B. (2011). Principles and practice of structural equation modeling. New York:
Guilford Press.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical
data. Biometrics, 33, 159174.
McFarland, K., & Anderson, J. (1980). Factor stability of the Edinburgh Handedness
Inventory as a function of testretest performance, age and sex. British Journal of
Psychology, 71, 135142. doi:10.1111/j.2044-8295.1980.tb02739.x
Milenkovic, S., & Dragovic, M. (2012). Modification of the Edinburgh Handedness
Inventory: A replication study. Laterality: Asymmetries of Body, Brain and Cognition,
Advance online publication. doi:10.1080/1357650X.2012.683196
Muthén, L. K., & Muthén, B. O. (2008). Mplus: User’s guide 5.1. Los Angeles: Muthén &
Muthén.
Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh inventory.
Neuropsychologia, 9, 97113. doi:10.1016/0028-3932(71)90067-4
Raykov, T. (1997). Estimation of composite reliability for congeneric measures. Applied
Psychological Measurement, 21, 173184. doi:10.1177/01466216970212006
Veale, J. F., Clarke, D. E., & Lomax, T. C. (2010). Biological and psychosocial correlates of
adult gender-variant identities: New findings. Personality and Individual Differences,
49, 252257. doi:10.1016/j.paid.2010.03.045
Williams, J. S. (1978). A definition for the common factor model and the elimination of
problems of factor score indeterminacy. Psychometrika, 43, 293306.
doi:10.1007/BF02293640
Williams, S. M. (1986). Factor analysis of the Edinburgh Handedness Inventory. Cortex, 22,
325326.
Edinburgh Handedness InventoryShort Form
15
APPENDIX
Edinburgh Handedness Inventory - Short Form
Please indicate your preferences in the use of hands in the following activities or
objects:
Always
right
Usually
right
Both
equally
Usually
left
Always
left
Writing
Throwing
Toothbrush
Spoon
Scoring:
For each item: Always right = 100; Usually right = 50; Both equally = 0; Usually left = -50;
Always
left = -100
To calculate the Laterality Quotient add the scores for the four items in the scale and divide
this by four:
Writing score
Throwing score
Toothbrush score
Spoon score
Total
Total ÷ 4 (Laterality Quotient)
Classification:
Laterality Quotient score:
Left handers
-100 to -61
Mixed handers
-60 to 60
Right handers
61 to 100
... The participants in this study were recruited through convenience sampling from a university population. The handedness of the participants was evaluated using the short version of the self-report Edinburgh handedness inventory [19]. Only right-handed participants were included, because there are differences in the cortical organization and motor performance between right-handed and left-handed people [20,21]. ...
Article
Full-text available
The Finger Tapping Test (FTT) is a classical neuropsychological test that assesses motor functioning, and recently it has been employed using smartphones. For classical protocols, it has been observed that sex and handedness influence the performance during the test. By assessing the influence of sex and handedness on the test, it is possible to adjust the performance measurements to ensure the validity of test results and avoid sex- and handedness-related bias. The present study aimed to evaluate the influence of sex and handedness on smartphone-based FTT performance. We developed an Android application for the FTT and recruited 40 males and 40 females to carry out three spatial designs on it (protocols I, II, and III). Participants' performance was measured using the global, temporal, and spatial parameters of the FTT. We observed that for the performance in protocol I, handedness had a significant influence on global and temporal variables, while the interaction between handedness and sex had a greater influence on spatial variables. For protocols II and III, we observed that handedness had a significant influence on global, temporal, and spatial variables compared to the other factors. We concluded that the smartphone-based test is partly influenced by handedness and sex, and in clinical implications, these factors should be considered during the evaluation of the smartphone-based FTT.
... To investigate this, we undertook a controlled experiment with 234 participants who had lived the majority of their lives in either the northern or southern hemisphere and made sure to balance the sample by handedness so that we had similar numbers of left-and right-handed participants from each hemisphere. Handedness was further confirmed using the Edinburgh Handedness Inventory Short Form (Veale, 2014). The sample was also balanced by gender. ...
... Twenty patients affected by chronic pain (7 males and 13 females, mean age ± SD: 62.9 ± 9.9) and 20 age-matched healthy subjects (12 men and 8 women, mean age ± SD: 59.6 ± 15.8) were recruited from the outpatient Clinic for Chronic Pain at the Department of Human Neurosciences, Sapienza University of Rome. All participants were right-handed, according to the Edinburgh Handedness Inventory (Veale, 2014). Diagnosis of chronic pain was based on clinical criteria with the aid of standardized clinical scales (Scholz et al., 2019). ...
Article
Objective: Chronic pain may lead to functional changes in several brain regions, including the primary motor cortex (M1). Our neurophysiological study aimed to probe M1 plasticity, through a non-invasive transcranial magnetic stimulation protocol, in a cohort of patients with chronic pain. Methods: Twenty patients with chronic pain (age ± SD: 62.9 ± 9.9) and 20 age- and sex-matched healthy controls (age ± SD: 59.6 ± 15.8) were recruited. Standardized scales were used for the evaluation of pain severity. Neurophysiological measures included laser-evoked potentials (LEPs) and motor-evoked potentials (MEPs) collected at baseline and over 60 minutes following a standardized Laser-paired associative stimulation (Laser-PAS) protocol. Results: LEPs and MEPs were comparable in patients with chronic pain and controls. The pain threshold was lower in patients than in controls. Laser-PAS elicited decreased responses in patients with chronic pain. The response to Laser-PAS was similar in subgroups of patients with different chronic pain phenotypes. Conclusions: M1 plasticity, as tested by Laser-PAS, is altered in patients with chronic pain, possibly reflecting abnormal pain-motor integration processes. Significance: Chronic pain is associated with a disorder of M1 plasticity raising from abnormal pain-motor integration.
... 529 SD=4.9), and all participants reported having normal or corrected-to-normal vision.530 Participants' handedness was assessed via the revised short form of the Edinburgh 531Handedness Inventory(Veale, 2014), with participants only permitted to take part if they 532 scored a laterality quotient of +40 indicating right handedness. Exclusion criteria for 533 participants were determined prior to data collection as follows: (1) if 10 or more channels 534 ...
Preprint
Full-text available
Do early effects of predictability in visual word recognition reflect prediction error? Electrophysiological research investigating word processing has demonstrated predictability effects in the N1, or first negative component of the event-related potential (ERP). However, findings regarding the magnitude of effects and potential interactions of predictability with lexical variables have been inconsistent. Moreover, past studies have typically used categorical designs with relatively small samples and relied on by-participant analyses. Nevertheless, reports have generally shown that predicted words elicit less negative-going (i.e., lower amplitude) N1s, a pattern consistent with a simple predictive coding account. In our preregistered study, we tested this account via the interaction between prediction magnitude and certainty. A picture-word verification paradigm was implemented in which pictures were followed by tightly matched picture-congruent or picture-incongruent written nouns. The predictability of target (picture-congruent) nouns was manipulated continuously based on norms of association between a picture and its name. ERPs from 68 participants revealed a pattern of effects opposite to that expected under a simple predictive coding framework.
... Evidence suggests that cerebral hemispheric dominance (indicated by handedness) can impact cognitive performance. In particular, since handedness may determine an advantage on spatial tasks and psychophysics (Somers et al., 2015), participants were assessed for handedness using the Edinburgh Handedness Inventory (EHI, Veale, 2014) and counterbalanced across the control and experimental groups (each contained one left-handed participant). ...
Article
Full-text available
Functional near-infrared spectroscopy (fNIRS) relies on near-infrared (NIR) light for changes in tissue oxygenation. For decades, this technique has been used in neuroscience to measure cortical activity. However, recent research suggests that NIR light directed to neural populations can modulate their activity through "photobiomodulation" (PBM). Yet, fNIRS is being used exclusively as a measurement tool. By adopting cognitive tests sensitive to prefrontal functioning, we show that a 'classical' fNIRS device, placed in correspondence of the prefrontal cortices of healthy participants, induces faster RTs and better accuracy in some of the indexes considered. A well-matched control group, wearing the same but inactive device, did not show any improvement. Hence, our findings indicate that the 'standard' use of fNIRS devices generates PBM impacting cognition. The neuromodulatory power intrinsic in that technique has been so far completely overlooked, and future studies will need to take this into account.
Preprint
Full-text available
Our aim was to study the processes involved in the spatial coding of the body during actions eliciting multiple effects simultaneously. We specifically aim to challenge the intentional-based account, which proposes that the effects used to code responses are those deemed relevant to the agent's goal. Accordingly, we used a Simon task, which is widely recognized as a suitable method for investigating the spatial coding of responses, combined with a setup inducing a multimodal discrepancy between visual and tactile/proprioceptive effects, known to be crucial for body schema construction and action control. To be more precise, the setup allowed to horizontally reverse the visual effects of the hands compared to the tactile/proprioceptive effects (e.g., the right hand was seen as being on the left). In Experiment 1, the participants' task emphasizing the relevance of visual effects, whereas in Experiment 2 it emphasized the tactile/proprioceptive effects. In Experiment 3, the visual effects were no longer reversed. We observed a Simon effect based on the location of the tactile/proprioceptive effects in Experiments 2 and 3. However, in Experiment 1 the Simon effect was partially driven by the location of the visual effects. Altogether, these findings provide support for the notion that the intention of the agent plays a prominent role in the representation of the body during action. This work also opens a promising avenue of research linking action and body representations.
Article
Motor imagery based brain-computer interfaces (MI-BCIs) have excellent application prospects in motor enhancement and rehabilitation. However, MI-induced electroencephalogram features applied to MI-BCI usually vary from person to person. This study aimed to investigate whether the motor ability of the individual upper limbs was associated with these features, which helps understand the causes of inter-subject variability. We focused on the behavioral and psychological factors reflecting motor abilities. We first obtained the behavioral scale scores from Edinburgh Handedness Questionnaire, Maximum Grip Strength Test, and Purdue Pegboard Test assessments to evaluate the motor execution ability. We also required the subjects to complete the psychological Movement Imagery Questionnaire-3 estimate, representing MI ability. Then we recorded EEG signals from all twenty-two subjects during MI tasks. Pearson correlation coefficient and stepwise regression were used to analyze the relationships between MI-induced relative event-related desynchronization (rERD) patterns and motor abilities. Both Purdue Pegboard Test and Movement Imagery Questionnaire-3 scores had significant correlations with MI-induced neural oscillation patterns. Notably, the Purdue Pegboard Test of the left hand had the most significant correlation with the alpha rERD. The results of stepwise multiple regression analysis showed that the Purdue Pegboard Test and Movement Imagery Questionnaire-3 could best predict the MI-induced rERD. The results demonstrate that hand dexterity and fine motor coordination are significantly related to MI-induced neural activities. In addition, the method of imagining is also relevant to MI features. Therefore, this study is meaningful for understanding individual differences and the design of user-centered MI-BCI.
Article
Prism adaptation (PA) is both a visuomotor learning task and potential treatment for spatial neglect after stroke. While PA's aftereffects can improve neglect symptoms, therapeutic benefits vary across individuals, possibly due to differences in neglect subtypes. Neglect symptoms can be described along an information processing pathway, yielding perceptual (input) and premotor (output) neglect subtypes. There is some evidence that PA mainly benefits persons with premotor neglect. We investigated whether PA modulates the premotor stage of information processing by examining whether PA could induce a premotor bias in healthy adults. We measured perceptual and premotor biases using a speeded reach task that compares the initiation time of leftward and rightward reaches to lateralized targets from different hand start positions. Using a randomized mixed experimental design, 30 right-handed healthy adults completed this speeded reach task before and after either left-shifting (n = 15) or right-shifting (n = 15) PA. As hypothesized, left-shifting PA speeded initiation time specifically for reaches in the rightward direction, regardless of target location (p = .02, ηp2 = .18), suggesting that PA induced a premotor bias in the direction of the prism aftereffect. These findings have implications for PA's underlying mechanisms, which can inform visuomotor learning theories and PA's use as a treatment for spatial neglect.
Article
Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique used in neurorehabilitation to enhance motor training. However, its benefits to motor training can be difficult to reproduce across research studies. It is possible that the observed benefits of tDCS are not directly related to the intervention itself but rather to the brain-mind responses elicited by the treatment context, commonly known as a placebo effect. This study investigated the presence of a placebo effect of tDCS on motor training and explored potential underlying factors. Sixty-eight participants who were right-handed were randomly assigned to active tDCS, sham tDCS, or a no-stimulation control group. Double-blind active or sham tDCS was applied to the right primary motor cortex, while the unblinded control group received no stimulation. All participants completed 30 training trials of a functional upper-extremity motor task. Participants' beliefs of tDCS, along with their prior knowledge of tDCS, were also collected. There was no significant difference in the amount of improvement on the motor task between the active and sham tDCS groups; however, both active and sham tDCS groups improved more than the control group, indicating a placebo effect. More motor task improvement was also associated with higher beliefs of tDCS (regardless of whether active or sham tDCS was received). This demonstrates a measurable placebo effect of tDCS on motor training, driven at least in part by treatment expectations or beliefs. Future tDCS studies should control for beliefs and other placebo-related factors.
Article
Full-text available
Researchers commonly compare means and other statistics across groups with little concern for whether the measure possesses strong factorial invariance (i.e., equal factor loadings and intercepts/thresholds). When this assumption is violated, inaccurate inferences associated with statistical and practical significance can occur. This manuscript emphasizes the importance of testing for measurement invariance (MI) and provides guidance when conducting these tests. Topics discussed are potential causes of noninvariant items, the difference between measurement bias and invariance, remedies for noninvariant measures, and considerations associated with model estimation. Using a sample of 491 teachers, a demonstration is also provided that evaluates whether a newly constructed behavior and instructional management scale is invariant across elementary and middle school teachers. Analyses revealed that the results differ slightly based on the estimation method utilized although these differences did not greatly influence the latent factor mean difference conclusions. Additional implications and considerations related to invariance testing are discussed.
Article
Let Ω be a domain in ℝ n whose boundary is C 1 if n≥3 or C 1,β if n=2. We consider a magnetic Schrödinger operator L W,q in Ω and show how to recover the boundary values of the tangential component of the vector potential W from the Dirichlet to Neumann map for L W,q . We also consider a steady state heat equation with convection term Δ+2W·∇ and recover the boundary values of the convection term W from the Dirichlet to Neumann map. Our method is constructive and gives a stability result at the boundary.
Article
A structural equation model is described that permits estimation of the reliability index and coefficient of a composite test for congeneric measures. The method is also helpful in exploring the factorial structure of an item set, and its use in scale reliability estimation and development is illustrated. The modeling. estimator of composite reliability it yields does not possess the general underestimation property of Cronbach's coefficient a.