On some objective and subjective aspects of discriminating between correlated and independent patterns and their relationship to Locus of Control and depression

ABSTRACT

Acknowledging the fact that there seems to be a well-established connection between depression and Locus of Control, we propose that Locus of Control is at least partly determined by one’s objective ability to discriminate between random and non-random stimuli and events. Further, we claim that this ability is indirectly related to depression through Locus of Control which serves as a mediator. In addition, we hypothesise that a subjective threshold, important in discriminating between random and correlated stimuli, is directly related to both depression and Locus of Control. Results from a study with 128 subjects indicate that the majority of our conjectures are supported by the data, the exception being that while the subjective threshold is significantly related to depression it appears to have no tangible influence on Locus of Control.

KEYWORDS:

• Random patterns
• Locus of Control
• depression

Acknowledgements

We would like to thank Velina Hristova for recruiting the majority of our test subjects. We are also thankful to two anonymous reviewers for providing us with several references which are highly relevant to our research project.

Disclosure statement

No potential conflict of interest was reported by the author(s).

ORCID

Svetoslav Bliznashki http://orcid.org/0000-0002-7232-8772

Notes

1 See for example “Emergent Techniques for Assessment of Visual Performance” by National Research Council (US) Committee on Vision, Washington (DC): National Academies Press (US), 1985; https://www.ncbi.nlm.nih.gov/pubmed/25032463

2 The only exception was Age which correlated negatively with d’ (r = −0.26, p=0.006). However Age didn’t correlate significantly with the other measures of interest and its inclusion in the path analyses reported below made no difference whatsoever with respect to the reported results. Also, Age was not involved in any interesting mediation and/or moderation effects.

3 As stated above, these results are for the subjects showing positive d’s only. The results, however, remain practically the same if all 128 subjects are included in the analyses. The d’ and Correct values in Table 1 are slightly lower but they retain their significant difference from the chance level of success; the remaining results are practically identical to the ones reported in Table 1 when calculated for the whole dataset.

4 We should note that several regression and path analyses indicated that the interaction between d’ and C had no significant direct and/or indirect effects on LoC and BDI and hence the interaction term was not included in the analyses reported below.

5 All path analyses were performed by the lavaan software for the R open source environment (Rosseel, 2012). BDI showed a significant deviation from the normal distribution based on a Kolmogorov-Smirnoff test (p<0.05) which necessitated bootstrapping of the standard errors for the path coefficients as well as for the overall model fit implemented within the same software.

6 We should point out that Clarke (2004) uses Levenson’s (1981) three-factor scale, IPC, in order to measure LoC and his Neuroticism’s mediation effect is observed primarily with respect to the Chance LoC sub scale.