Hazardous Attitudes Scale - New Version


This instrument is derived from a measure developed by Holt and associates (Holt et al., 1991; Holt, Boehm-Davis, Amendola, & Sweeney, 1994) at George Mason University. They designed a new instrument consisting of simple declarative statements with a Likert-type response scale that they believed would be superior to the old ipsative measurement scale. However, because of the difficulty associated with evaluating the instrument with pilots, they developed an equivalent form suitable for drivers and administered that version to a sample of 238 undergraduate students. Factor analysis showed that the instrument had factors that generally corresponded to four of the factors (macho, impulsivity, antiauthority, and resignation) purported to be measured by the FAA’s instrument. Additionally, a factor representing confidence or competence in driving was found.

Holt et al. (1991; 1994) found that factor scores from the new instrument were significantly correlated with a number of criteria of interest including, seat belt usage, drinking and driving, moving violations, and accidents and incidents. Although Holt et al. appropriately limited their discussion of the results to drivers, their findings raise the question of whether similar outcomes could be obtained with pilots.

Subsequently, Hunter (2005) administered this and several other attitude scales to large samples of pilots. The data presented here are from Hunter (2005).

Number of Items: 88
Format: Likert scale, with responses from Strongly Agree (5) to Strongly Disagree (1).


The instrument, and the subscales are scored by the simple summation of the keyed items. (Note the reverse-keyed items below.)

Factor1 = q1 + q2 + q3 + q4 + q5 + q8 + q9 + q10 + q11 + q12 + q70 + q72 + q73.

Factor2 = q57 + q53 + q54 + q49 + q47 + q56 + q55 + q59 + q45 + q51 + q46 + q48 + q52.

Factor3 = q61 + q76 + q88 + q86 + (6 - q83) + (6 - q84) + q87 + (6 - q79) + (6 - q77) + q80 + q81.

Factor4 = q24 + q20 + q21 + q22 + q30 + q23 + q28 + q26 + q18 + q13 + (6 - q19) + q31.

Factor5 = q33 + q38 + q82 + q35 + (6 - q16) + q43 + q34 + q37.

Factor6 = q69 + q64 + q78 + q60 + q66 + q51 + q68.

NOTE: The scoring given above is for the 88-item version of the scale. This is the version that was used in the studies cited here. An abbreviated (30-item) version of the scale is offered on this web site. The abbreviated version is intended to develop self-awareness of hazardous attitudes among pilots. If you plan to use this scale in research, I recommend you use the original 88-item version because it will provide better reliability and comparability to previous studies.

However, if you want to use the 30-item version, then the scoring for that scale is:

Macho = Q5 + Q7 + Q12 + Q22 + Q30
Resignation = Q14 + Q15 + Q17 + Q21 + Q25
Antiauthority = Q13 + Q20 + Q24 + Q26 + Q29
Worry = Q2 + Q3 + Q8 + Q10 + Q16 
Impulsive = Q6 + Q9 + Q11 + Q18 + Q23
SelfConfidence = Q1 + Q4 + Q19 + Q27 + Q28

[Question numbers refer to the 30-item version]

Means and SDs for the sub-scales based on 30-item version (from a sample of approximately 6,000 pilots) are:























Factor Composition:

The analysis of the New–HAS responses yielded an interpretable six-factor solution that accounted for 36% of the variance. These factors generally corresponded to those previously identified by Holt et al.
(1991), and included Macho, Resignation, Antiauthority, Worry/Anxiety,
Impulsivity, and Self-Confidence.

See the Normative Information, below.

Construct Validity:

Because there is no single instrument that is accepted as the standard of measurement in this area, several measures were used to evaluate the construct validity of the attitude scales, each representing slightly different convergent external constructs or measures of interest. These measures included Zuckerman’s (1994) TAS, the ASLOC (Aviation Safety Locus of Control), measures of pilots’ risk perception and risk tolerance, a measure of pilots’ situational judgment, and a measure of their involvement in hazardous aviation events including accidents.

Eight of the 60 correlations for the New–HAS were statistically significant ; however, these correlations were generally substantially larger than those obtained by the Old–HAS. In particular, patterns of substantial, significant correlations were found for the Antiauthority scale and the Worry/Anxiety scale. The Antiauthority scale was significantly correlated with the TAS Scale and with the LOC scales. The Worry/Anxiety scale was significantly correlated with both of the risk perception measures and with one of the risk tolerance measures. As would be expected, the Resignation scale was correlated significantly with the LOC Externality scale. In addition, the Impulsivity scale was negatively correlated with the LOC Internality scale, suggesting that those pilots who believed themselves to be more in control of the situation were less likely to agree with statements reflecting impulsivity.


Normative Information:

From Hunter, 2005 M SD n of items Coeff Alpha N
Macho 34.84 7.69 15 .86 191
Resignation/External 23.85 6.10 13 .82 188
Antiauthority 23.99 5.60 11 .82 190
Worry/Anxiety 38.22 6.67 12 .83 191
Impulsive 19.68 4.48 8 .74 190
Self-Confidence 23.46 3.57 7 .59 190



Holt, R. W., Boehm-Davis, D. A., Amendola, K. L.,& Sweeney, M. M. (1994). The influence of thought patterns on driving performance. Unpublished manuscript, George Mason University, Fairfax, VA.

Holt, R. W., Boehm-Davis, D. A., Fitzgerald, K. A., Matyuf, M. M., Baughman, W. A.,&Littman, D. C. (1991). Behavioral validation of a hazardous thought pattern instrument. In Proceedings of the Human Factors Society 35th Annual Meeting (pp. 77–81). Santa Monica, CA: Human Factors Society.

Hunter, D.R. (2005). Measurement of hazardous attitudes among pilots. International Journal of Aviation Psychology, 15, 23-43.

 Download the scale (Word format)