Desktop version

Home arrow Philosophy

Content validity

This refers to the instruments ability to cover the full domain of the underlying concept. There is considerable debate about this at the moment. Whilst it is clearly possible to write a very short test that has excellent reliability, the usefulness of such a test can be questionable. There are a number of very short quick tests available, but because of their limited number of items they have some difficulty providing a useful differentiation between individuals. Conversely, if you make a test too long, ensuring it covers every possibly aspect and item, its usability declines rapidly and its psychometric integrity is severely challenged. We have therefore tried to design a test that has the complexity to reflect the nuances of mental toughness but short enough for the test to be both useful and useable.


Construct validity

This is a rather complex idea. It relates to the question of whether or not mental toughness is actually a valid concept. Does it exist as a discrete and useful entity? One approach is to ensure that the concept being measured relates to other concepts that are similar in nature. The results of our initial attempt to do this, which was carried out in the development phase of the test, are reported in Table 2.

These correlations are both in the expected direction and are obviously explainable. For example, mental toughness should be associated with lower levels of trait anxiety and it can be seen that it is.

The validation of the questionnaire and the model has continued apace since its initial development. For example, Horsburgh and colleagues (2009) found evidence that mental toughness was significantly related to the big five personality factors of extraversion, neuroticism (negatively), agreeableness, openness to experience, and conscientiousness.

In developing the MTQ48, Clough, Earle and Sewell (2002) found that mental toughness consisted of six components or factors. These factors were, first, challenge, the extent to which individuals see problems as opportunities for self-development. Second, commitment, which

Table 2. Correlations table for MTQ48 and various personality scales.

Life orientation test

Satisfaction with life scale

Self-esteem scale

Self-efficacy scale

State trait anxiety inventory

Overall MT

0.48**

0.56**

0.42*

0.68**

−0.57**

Challenge

0.39*

0.59**

0.45*

0.66**

−0.54**

Commitment

0.45*

0.52**

0.40*

0.69**

−0.59**

Control

0.49**

0.55**

0.41*

0.64**

−0.61**

Control: life

0.53**

0.59**

0.49**

0.66**

−0.63**

Control:

0.46*

0.56**

0.34*

0.59**

−0.61**

emotions Confidence

0.47*

0.50**

0.39*

0.70**

−0.58**

Confidence:

0.49**

0.49**

0.45*

0.74**

−0.60**

in abilities Confidence:

0.41*

0.56**

0.37*

0.69**

−0.61**

interpersonal

Note: ** p < 0.01, * p < 0.05. reflects deep involvement with whatever one is doing; third, emotional control, the ability to keep anxieties in check and not reveal emotions to others. Fourth, life control, concerning a belief in being influential and not controlled by others. Fifth, confidence in abilities reflects belief in individual qualities with less dependence on external support, and sixth, interpersonal confidence, this is about being assertive and less likely to be intimidated in social contexts.

The second approach to establishing construct validity is more statistically based. Since the identification of the four C's model, independent researchers have tested the factor structure of the MTQ48 using a statistical procedure called confirmatory factor analysis. Horsburgh, Schermer, Veselka and Vernon (2009) provided support for the factor structure proposed by Clough and colleagues (2002) using a North American sample. There has since then been some vigorous academic debate around the structure of the model. We therefore decided to carry out a very large study to try and resolve the debate. A total of 8,207 participants took part in this study. The participants in this study consisted of 4,342 senior managers, 1,440 lower and middle managers, 1,004 clerical/administrative workers, 442 athletes, and 978 students. Model fit was assessed using confirmatory factor analysis (CFA) and exploratory structural equation modeling, in addition to the robust maximum likelihood estimator. Overall, the results support the factorial validity of the MTQ48 and indicate that the MTQ48 is a robust psychometric measure of mental toughness. It also supports the model itself.

Criterion validity

Finally there is criterion related validity. We feel this is the both the core and the key to the MTQ48. It is a measure of whether or not a score on the MTQ48 is associated with an external measure. Much of our work has concentrated on performance measures. This type of evidence is really at the core of this book. A number of case studies and published studies described in the other chapters reveal an impressive foundation of criterion validity. We believe that the test and model are truly useful and can help enhance both performance and provide an explanation of behaviours in a number of domains.

Summary

The MTQ48 and its associated model have proved themselves over a decade of research. The MTQ48 provides a reliable measure of mental
toughness. The model of mental toughness it measures has clear theoretical roots and appears to offer a solid understanding of what mental toughness means. Although there are many models and theories of toughness, we feel ours has a proven track record. It is certainly not the only way of conceptualising this interesting concept, but we feel it offers a parsimonious and useful way forward.


 
< Prev   CONTENTS   Next >

Related topics