Indian Journal of Occupational and Environmental Medicine   Official publication of Indian Association of  0ccupational  Health  
 Print this page Email this page   Small font sizeDefault font sizeIncrease font size
 Users Online:485

  IAOH | Subscription | e-Alerts | Feedback | Login 

Home About us Current Issue Archives Search Instructions
  Search
 
  
 
    Similar in PUBMED
     Search Pubmed for
     Search in Google Scholar for
    Article in PDF (306 KB)
    Citation Manager
    Access Statistics
    Reader Comments
    Email Alert *
    Add to My List *
* Registration required (free)  


   References

 Article Access Statistics
    Viewed335    
    Printed3    
    Emailed0    
    PDF Downloaded29    
    Comments [Add]    

Recommend this journal

 


 
  Table of Contents 
LETTER TO THE EDITOR
Year : 2016  |  Volume : 20  |  Issue : 2  |  Page : 121
 

Post hoc tests in analysis of variance


Department of Community Medicine, Indira Gandhi Medical College, Shila, Himachal Pradesh, India

Date of Web Publication4-Jan-2017

Correspondence Address:
Anjali Mahajan
Department of Community Medicine, IG Medical College, Shimla - 171 001, Himachal Pradesh
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/0019-5278.197552

Rights and Permissions

 



How to cite this article:
Mahajan A. Post hoc tests in analysis of variance. Indian J Occup Environ Med 2016;20:121

How to cite this URL:
Mahajan A. Post hoc tests in analysis of variance. Indian J Occup Environ Med [serial online] 2016 [cited 2017 Apr 28];20:121. Available from: http://www.ijoem.com/text.asp?2016/20/2/121/197552


Dear Sir,

Apropos the article "Sources of occupational stress in the police personnel of North India: An exploratory study" by Shweta Singh and Sujita Kumar Kar published in your journal as an original article. [1] I would like to congratulate the authors for bringing out this important issue, which has evoked international concern. Stress at work place is experienced by all. However, the magnitude of the same may vary in different professions, and furthermore it depends upon one's personal situations and conditions at workplace, as highlighted in the present study.

I would like to raise an important concern which was not addressed in the present study regarding the analysis of attributes of stress in various groups of police personnel. A one-way analysis of variance was conducted to assess if there were differences in terms of various sources of stress in inspectors, constables, and officers. The independent categorical variable was three categories of police personnel and the dependent continuous variable was means of overall and various other dimensions of stress levels. Analysis of variance (ANOVA) tests the nonspecific null hypothesis that all three population means are equal. This nonspecific null hypothesis is sometimes called the omnibus null hypothesis. When the omnibus null hypothesis is rejected, the conclusion is that at least one population mean is different from at least one other mean. [2] Because the null hypothesis is rejected if any pair of means is unequal, where the significant difference lies needs to be worked out which was not done in the present study. However, since the ANOVA does not reveal which means are different from which, it offers less specific information. The point is that after the overall analysis (often called an omnibus F test) we are left in an ambiguous situation. To eliminate the ambiguity regarding the results of the study, it is necessary to compute what are called post hoc or subsequent analyses (also called protected t-tests because they always involve two means and have a reduced chance of making a Type I error). After completing the post hoc analyses, we would know exactly which means differed, and hence we would be in a position to completely interpret the outcome of the study. In the present study, according to the authors, inspectors had significantly higher stress in the domains of private life, organizational structure, and interpersonal stressors. Officers had the highest stress in the domains of management of people and role ambiguity. However, constables were also experiencing the same in the domains of environmental stress, work pressure, and travelling. These findings are based on the observed difference in means and therefore provide only the numerical summary. It would have been preferable if the authors had run post hoc comparisons, which would have provided us with statistical summary so as to know whether difference between groups were large enough to say that there are statistical difference between different groups of the police personnel with reference to different attributes of stress.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

 
  References Top

1.
Singh S, Kar SK. Sources of occupational stress in the police personnel of North India: An exploratory study. Indian J Occup Environ Med 2015;19:56-60.  Back to cited text no. 1
[PUBMED]  Medknow Journal  
2.
Maxwell SE, Delaney HD. Looking for Group Differences with a One-Way Analysis of Variance-Effects of Planning Time. In: Designing Experiments and Analyzing Data: A model comparison perspective, 2 nd ed. Lawrence Erlbaum Associates; 2004. p. 269.  Back to cited text no. 2
    




 

Top
Print this article  Email this article