Laterality: Exploring the Enigma of Left-Handedness

11.2.

Chapter 1

Everybody’s Right, So What’s Left?

Abstract

Handedness is measured by tests of hand preference, the preferred hand used for one-handed activities like writing, and hand performance, skill, and timing differences between the hands for activities like placing pegs in holes or finger tapping. There are many hand preference questionnaires but issues of questionnaire length, scoring procedure, and classification system make the identification of left-handers problematic. Left-handers are more likely to report the use of the right hand for some activities than right-handers to report the use of the left hand. For this reason, left-handers are at higher risk of misclassification when questionnaire scores are used to form right- and left-handed groups. Left-handers may also show lesser degrees of asymmetry in skilled hand performance when the preferred hand actions are compared to those of the nonpreferred hand. The classification of individuals into right- and left-handed groups is fluid but there are recommended best practices for the measurement of handedness.

Keywords

Hand preference; hand performance; hand preference questionnaires; hand skill; movement lateralization; handedness

It may seem strange to start a book about left-handedness with the question, What’s left?, but the measurement of left-handedness is elusive. Identifying left-handers can be controversial, so it is important to address this question at the outset. The side of the writing hand is the most frequently used criterion for calling someone right- or left-handed. Many researchers think this is an adequate basis for defining handedness. However, others claim that a single gold standard for determining a person’s handedness does not exist [1].

Historically, scientists have measured handedness in two ways. They ask people to describe their hand use when performing tasks where only one hand can be used, such as when one writes or draws. This method measures hand preference. The second approach requires that people perform a number of tasks such as placing pegs in small holes as rapidly as possible. The time to complete the task for both the right and the left hands is computed and compared. This method assesses hand performance. Table 1.1 lists examples of hand preference and performance behaviors. Hand preference measures examine one-handed activities and are classified as showing either strong or weak lateralization toward one hand. A strongly lateralized behavior is done consistently with the same hand, right, or left. A weakly lateralized behavior can be done, at least sometimes, with either the right or the left hand. Hand performance tasks measure both right- and left-handed action and then the two results are compared.

Table 1.1

Examples of Handedness Measures

Hand Preference

The most common method for determining the side of handedness is to measure hand preference. Hand preference is typically assessed through questionnaire responses rather than having people actually perform one-handed behaviors. Occasionally, I am asked why handedness researchers have concentrated on developing questionnaires and why they depend so heavily on survey methods. My response is to point to the human population handedness asymmetry estimated worldwide at 85–90% right-handers versus 10–15% left-handers. Left-handedness is rare, so large samples are tested to find a reasonably sized group of people who use the left hand at least for some activities. Hand preference questionnaires can be administered to groups, sent in the mail and posted on the internet allowing researchers to screen hundreds and even thousands of people to find the left-handed minority [2].

Table 1.2 compares the specifics of some of the major hand preference questionnaires developed over the last 90 years. Handedness researchers try to insure that their questionnaires have validity. In other words, does the questionnaire measure what it claims to measure? The hand preference behaviors listed in Table 1.1 have face validity. These behaviors reflect a general understanding of one-handed preferences and include activities people use to classify their own hand preference. However, researchers also want to establish the behavioral validity of questionnaire responses by comparing the way people answer the questions with their actual hand use. For example, I can ask a person to write their name and compare the hand I observe to complete this task with the hand the person reports using when asked this question on a questionnaire. If the questionnaire has good behavioral validity, actual hand use is the same as the questionnaire response [3,4].

Table 1.2

Examples of Hand Preference Questionnaire Formats

Source: Ref. [3].

Many different hand preference questionnaires have been developed over the years. Researchers analyze them to determine if some are better than others in measuring true hand preference. Also, they want to identify problems with the questionnaire to improve both its content and its administration. These efforts have led to several controversies that demonstrate the difficulty in determining just who has right- or who has left-hand preference [5].

Hand Preference Questionnaires: How Many Questions?

Hand preference questionnaires vary in length as shown by the examples in Table 1.2. Many researchers study handedness types because they want to explore the relationship between handedness and brain organization, particularly language lateralization to the right versus the left hemisphere. These researchers want an accurate assessment of the side of hand preference but they also want a brief questionnaire, 10 items or fewer, which can be administered easily to many different types of study participants. Other researchers are focused on understanding hand preference itself and want to explore the various dimensions of strong and weak lateralization and two-handed behaviors where one hand is preferred. For example, a widely used 32-item hand preference questionnaire contains (i) strongly lateralized one-handed behaviors, such as hammering nails and writing, (ii) weakly lateralized one-handed items, such as picking up a book, and (iii) activities involving two hands where one hand takes the lead, such as in using an axe or batting a ball. Other two-handed activities (i) where one hand takes the lead, such as opening a jar lid while the other hand holds the jar or (ii) where one hand threads a needle while the other hand holds the needle may also be included. All of the questionnaires in Table 1.2 contain two-handed items but 60 questions may be too cumbersome. A generally useful hand preference questionnaire should contain enough items to cover both one-handed and two-handed activities. One suggestion is to use a questionnaire between 10 and 25 items. This length of questionnaire is easily administered and provides adequate information about the range of hand preference behaviors [6].

Hand Preference Questionnaires: How Are the Questions Scored?

Researchers use a variety of procedures to produce a numerical score representing a person’s hand preference pattern. Applying numbers to the answers to questionnaire items would be easy if respondents answered all the questions in the same way. However, diversity rather than consistency is the usual result when calculating questionnaire responses, and this is especially true for left-handers. The last three columns of Table 1.2 list the percentages of people classified as all right, all left, and mixed as determined by the seven questionnaires contained in the table. On average, 56% of respondents answered all the questions with right-hand use, 7% always answer with the left hand, while the remainder 37% answered the questions with a mixed pattern of right and left responses. One is about eight times more likely to find a consistent right-hander than a consistent left-hander. Also, hand preference is a continuous dimension both across and within individuals. The continuum ranges from consistent use of the right hand to consistent use of the left hand with many people falling between these two extremes.

Individuals vary in both their hand use and their responses across different items on a hand preference questionnaire. Only the oldest questionnaire in Table 1.2, published in 1927, uses a dichotomous scoring scheme of right versus left. More recent questionnaires incorporate a third category, either, or respondents are asked to rate themselves on scales ranging from always left to always right. A scoring scheme of −1, 0, +1 is used for the questionnaires with the three categories of left, either, right, respectively. The questionnaires with more elaborate categories may score all right responses as 5, mostly right responses as 4, either as 3, mostly left as 2, and all left responses as 1. The numbers assigned to each answer are summed across all the questionnaire items, and each person receives a hand preference score based on the sum or on the average of the sum.

There is one exception to these common scoring procedures as shown in row 4 of Table 1.2. The Edinburgh Handedness Inventory was developed by Oldfield in 1971. This inventory uses a ++and a +procedure to indicate the consistency of the preference for the left or the right hand. It is widely used with more than 15,000 citations in the scientific literature since its original publication. However, the dependence of researchers on this one questionnaire over the years has been called into question. First, respondents find the questionnaire format confusing, and only 47% of them complete the inventory correctly. Second, it has been modified so many times over the years that there are now many variants of the original inventory. These different variants produce diverse handedness classifications across studies even though researchers claim to be using the same questionnaire. Comparisons of handedness-related results among different studies are difficult, because researchers are not using the same original version of the questionnaire [7].

Table 1.3 is an example of a 10-item 3-response category questionnaire published in 2013 by Michael Nicholls and his associates. This instrument was developed after completing a detailed analysis of responses to items from two widely used questionnaires published in the 1970s (the Edinburgh Handedness Inventory was not used in this study). The researchers argued that it was time to collect new data and update these classic hand preference measures. The questionnaire responses result in a hand preference score that is the algebraic sum of the number of left, either, and right responses scored according to the scheme detailed in Table 1.3 [8].

Table 1.3

A 10-item Hand Preference Questionnaire with Three Response Categories

Table 1.4 is an example of another common approach to the measurement of hand preference. It is a 20-item 5-response category questionnaire I have used in studies of hand preference. I developed this questionnaire based on hand preference inventories published in the 1980s. The numbers assigned to each item are summed, and an average score is computed by dividing the number of items by 20. This procedure results in a hand preference score ranging from 1, left hand always, to 5, right hand always [9].

Table 1.4

A 20-item Hand Preference Questionnaire with Five Response Categories

The questionnaire is based on Ref. [9a].

Hand Preference Questionnaires: How Are Hand Preference Types Defined?

The questionnaires in Tables 1.3 and 1.4 result in hand preference scores that range along a continuous dimension. People use their right or their left hand for every activity listed, they are consistently handed, or they have a mixed pattern of hand use. A mixed pattern can either take the form of using the right hand for some activities and the left hand for others or being able to use either hand at will to perform a single task. It is rare to find someone who can write equally well with both the right and the left hand, so writing is considered a strongly lateralized behavior and a major predictor of overall hand preference side. Alternatively, several activities in Table 1.4, waving goodbye and picking up a glass of water, are weakly lateralized, and many respondents report being able to use either hand to perform these actions. The questionnaire in Table 1.3 is composed primarily of strongly lateralized items, while that in Table 1.4 contains a mixture of both strongly and weakly lateralized tasks.

A person can measure his or her hand preference by answering the items in both questionnaires and comparing the scores. To do a comparison, one must know how the hand preference types are defined once the final questionnaire score is computed. For example, the inventory in Table 1.3 results in scores ranging from −10, all left, to +10, all right. The researchers who developed this questionnaire suggest that if the final score falls between −10 and −5, the respondent is classified as left-handed, and if it falls between +5 and +10, the person is called right-handed. Individuals scoring between those cutoff points are called mixed-handed. However, this is just one approach to form hand preference categories. A researcher might use an alternate classification scheme if interested in the consistency of hand preference behaviors. Left-handers could be identified as only those with a score of −10, right-handers as +10 and everyone else in between is mixed-handed [10].

When I have used the questionnaire in Table 1.4, I have classified respondents with average scores of 1 and 2 as left, those with 4 and 5 as right, and those with 3 as mixed. In one study, I compared consistent and inconsistent left-handers. This distinction is made by comparing answers to questions about behaviors that require fine motor coordination, such as writing, to those behaviors that require more strength or whole arm movement, such as throwing. Approximately one-third of left-handers are inconsistent; they write with the left hand but throw with the right. Inconsistent left-handers may constitute a special subgroup of left-handers. I called scores of 1 and 2 as left and scores of 4 and 5 as right for both the writing and the throwing hand questions and compared the resulting combinations to derive the consistent and inconsistent groups [11].

Those who complete both Table 1.3 and Table 1.4 hand preference questionnaires should find that the hand preference scores agree when the suggested categorization schemes are used. If the score from Table 1.3 is between −5 and −10, left, the average score from Table 1.4 should be between 1 and 2, left. The results should be similar in the right-hand range. Respondents in the mixed ranges may find that the two questionnaires do not agree. The entries in Table 1.2 indicate that, on average, many fewer left-handers (7%) than right-handers (56%) respond consistently to hand preference questionnaire items. More left-handers than right-handers are likely to be classified as mixed-handed if one looks at the side of the writing hand versus other hand preference behaviors. Also, left-handers are more likely to be misclassified as right-handers when the hand preference questionnaire contains only a few items