Original Article

The Validity and Reliability of the Turkish Pet Attitude Scale

10.4274/meandros.galenos.2021.77045

  • Rahşan Çevik Akyıl
  • Seyhan Çitlik Sarıtaş
  • Ayşegül Kahraman
  • Musa Şamil Akyıl
  • Filiz Abacıgil

Received Date: 07.09.2021 Accepted Date: 29.10.2021 Meandros Med Dent J 2022;23(1):94-100

Objective:

This study tests the Turkish validity and reliability of the Pet Attitude Scale (PAS) and adapt it to Turkish society.

Materials and Methods:

The study population consisted of students in the departments of health at a university in Western Turkey. No sampling method was used; and the entire population was included in the study. Data were collected using a questionnaire on the participants’ sociodemographic characteristics and the PAS, which consisted of 18 items scored between 1 and 7 under three factors: love and interaction, pets in the home, and joy of pet ownership. The construct validity of the scale was evaluated by Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA). The reliability of the PAS was assessed by Cronbach’s alpha internal consistency coefficient, the item-total correlation, and the test-retest analysis.

Results:

The scale factors’ Cronbach’s alpha reliability coefficient ranged between 0.45 and 0.92, the items’ factor loads ranged between 0.51 and 0.85, and the total item correlation coefficients ranged between 0.31 and 0.77. The CFA supported the three-factor structure yielded by the EFA and showed good fit indices.

Conclusion:

The scale was reliable for Turkish society compared to the original PAS. The factor structure resembled the results of a reevaluation of PAS. Thus, it can be suggested that the PAS is valid and reliable for application in Turkish society.

Keywords: Pet attitude scale, Turkish version, validity, reliability

Introduction

Animal-assisted interventions are interventional programs that constitute integral parts of targeted treatment procedures. They are supportive methods for ameliorating physical, psychological, and social problems of individuals (1). Studies have shown that being with pets regulates the hemodynamic parameters such as heart rate, increases neurohormonal activity, and reduces pain (2-4).

Although the relationship between having a pet and psychological health could not be shown, most studies have reported that people with a pet have better mental health (4,5). Various studies have shown that supporting positive moods affects empathy and socialization, and reduces the sense of loneliness (6-9).

Pets affect the general health of weaker populations such as elderly people, children, and those with psychiatric disorders, and even our relationships with other people (10). It was found in the elderly that social and cardiovascular health was positively affected by the support of pets (11). It was also reported that support of pets was effective on metal, physical, and social health and wellness of children and that it was recommended to be applied for educational and health purposes in children with special needs (12). A study reported that a positive approach towards pets encourages people to approach other people more positively (13).

The literature includes few studies on people-animal relationships in Turkey (14,15). The number of pets in homes has significantly increased in Turkey in the last decade. In addition, many pets on the streets often lead to contact between humans and animals. Along with the increase in the number of house pets in urban areas, various opinions have also come forth in society regarding the relationship between humans, pets, and houses. Institutional and social awareness has also increased on animal rights, welfare, and even animal health (16). This suggests that this topic will become a popular and required research field in Turkey soon.

Various measurement tools exist to assess the quality of pet owners’ relationships (17-21). It was determined that the Pet Attitude Scale (PAS) was the commonly used measurement method (13,22-24) and the same scale using the form in English was applied to individuals in different cultures (25).

Bringing the scales on which knowledge had accumulated and recognized in international publications into Turkish language and using them shortens the time that an investigator would spend for preparing a new scale and provides the investigator ease of communication and information comparable to the results obtained in different communities. In addition, adapting the scales to Turkish might not facilitate researchers who do not have enough accumulation of knowledge and experience to develop a scale (26).

Römpke, while implementing the adaptation of PAS to the German language, stated that testing individuals with measurement tools in a language appropriate for their own culture would be more reliable, and more precise results might be gotten with better understanding the survey questions asked in their native language (27). 

The fact that no scale had been developed in Turkey on this subject was the determinative factor in planning our study. With the suggestion that it would be a preliminary reference work for initiation of research on human-animal relationships, this study aimed to test the validity and reliability of the PAS and to implement its adaptation to the Turkish society.


Materials and Methods

Aim

This study aims to adapt the PAS into Turkish and test its validity and reliability.

Study Design and Participants

This was a methodological study. Students who were receiving education and training in various health departments of a university located in Western Turkey constituted the study group.  A sample size of 300 participants was required to find a test’s factor structure (28). The research was completed with 304 students who filled the forms completely. For gathering research data, the questionnaire, which was prepared by the investigators and containing sociodemographic characteristics of students, and PAS were used.

Ethical Consideration

Permission was obtained via e-mail for the adaptation of the PAS, which was developed by Templer et al. (29), into Turkish. In addition, the approval of the Faculty of Dentistry, Clinical Research Ethics Committee was obtained with number 2018/026 to conduct the study. The participants were informed about the study and that their personal information would be kept confidential. Those who volunteered were included in the study.

Data Collection

The data was collected with the students in the departments of health at a university in Western Turkey on weekdays during the course breaks between May 14 and June 5, 2018. The polls were distributed to the students, who were asked to complete the questionnaires. A second interview was carried out with 77 of the students three weeks later. The data collection process took approximately 12 to 15 minutes for each student.

The Pet Attitude Scale

PAS was modified by Templer et al. (29) in 2011 and had 18 items as a 7-point Likert type scale. Each item was scored between 1 and 7. The scale had three subscales that indicated individuals’ attitudes towards pets, which were Love and Interaction (items 1, 2, 3, 5, 7, 8, 10, 11, 14, 16, 18), Pets in the Home (items 6, 9, 12, 13, 15) and Joy of Pet Ownership (items 4,17). The minimum and maximum scores were 18 and 126, respectively. The PAS was completed in 7 to 10 minutes. Templer et al. (29) found Cronbach’s α coefficient to be 0.92 for PAS (29).

Language Validity

First, the PAS was translated from English into Turkish to test language validity. Then, the Turkish version was back-translated into English by two linguists. The researchers reviewed the English back-translation by comparing it to the original scale. A preliminary study was conducted with 18 students using the PAS. No item was removed from PAS.

Statistical Analysis

Data were analyzed using the Cronbach’s α reliability coefficient, the Pearson Product-Moment Correlation Coefficient, factor analysis, the Barlett test, the Kaiser-Meyer Olkin test, correlations, numbers, and percentages. The p<0.05 was considered significant.


Results

The sociodemographic characteristics of the students are shown in Table 1. The average age of the students was 22. Of them, 62.8% were female, 99% were single, 72.4% studied nursing, 21.7% studied medicine, and 5.9% studied dentistry.

Validity

An EFA was applied using varimax rotation to determine the dimensional structure of the scale. The KMO value was used to test the suitability of the items for factor analysis, and the value obtained was determined to be 0.911. All values for the measure of sampling adequacy (MSA) were determined to be greater than 0.5 (all MSA >0.872). The Bartlett test for sphericity confirmed that the data was suitable for factor analysis (χ2=2778.40, df=153, p<0.001). In the first attempt, a 3-factor structure with an eigenvalue greater than 1 was obtained. The first factor consisted of the items 1, 2, 3, 5, 7, 8, 10, 11, 14, 16 and 18, whereas the second factor consisted of the items 6, 9, 12, 13, and 15 and the third factor consisted of the items 4 and 17. These three factors were determined to explain 57.4% of the total variance. The CFA performed for this structure revealed χ=467.614, df=132 (p<0.001), χ2/df=3.54, RMSEA=0.075, GFI=0.86, CFI=0.88, and IFI=0.89. The values obtained were determined to be outside the acceptable levels, even though close. Additionally, when the levels of internal consistency were analyzed, it was found to be 0.92 for the first factor, 0.76 for the second factor, and 0.45 for the third factor; the internal consistency was not at an acceptable level for the third factor. When the results of CFA and the problems related to the internal consistency level of the third factor were considered, it was concluded that such a 3-factor structure was not suitable, and the suitability of a 2-factor structure was tested.

The evaluation revealed that the first factor consisted of items 1, 2, 3, 5, 7, 8, 10, 11, 14, and 16, whereas the second factor consisted of items 4, 6, 9, 12, 13, 15, and 17. Therefore, the first factor was observed to consist of items related to positive behaviors, whereas the second factor consisted of items related to negative behaviors. The two factors were determined to explain 53.7% of the total variance. The factor loads of the items in the first factor were found to range from 0.569 to 0.883, whereas the factor loads of the items in the second factor from 0.592 to 0.832. CFA performed for this structure revealed χ2=279.891, df=128 (p<0.001), χ2/df=2.19, RMSEA=0.063, GFI=0.905, CFI=0.949, IFI=0.950 and SRMR=0.049. The results confirmed that this 2-factor structure had an acceptable level of suitability. The CFA Path Diagram of PAS-TR following the construction of the second CFA model was shown in Figure 1.

Reliability

Cronbach’s α was used to determine the internal consistency level of PAS-TR. The internal consistency level was found to be 0.925 for the first factor, 0.824 for the second factor, and 0.906 for the entire scale. When the corrected item-total correlation levels were analyzed, it was determined that the items with the highest level were the items 5 and 8 (0.772 and 0.733, respectively) and those with the lowest level were the items 9 and 12 (0.313 and 0.370, respectively).

The test-retest results of the PAS revealed that the intraclass correlation coefficient (ICC) level was 0.738 (good) for the first factor, 0.776 (excellent) for the second factor, and 0.773 (excellent) for the entire scale (30).


Discussion

In this study, PAS, which was modified by Templer et al. (29) in 2011, was adapted into Turkish. The psychometric features were tested on students in the departments of health in Turkey.

Validity

EFA and CFA were performed to determine the structural validity of PAS-TR. It was found that the scale was grouped under two factors in the Turkish version like the original scale and explained 53.7% of the total variance. A variance ratio ranging from 40% to 60% determined in the analysis of the scale has been considered sufficient (31). It has been shown in studies that items with a factor load below 0.30 must be not including during evaluation (32,33). It was determined in our study that the factor loads of PAS-TR ranging from 0.559-0.853 (Table 2); thus, no one the scale items were removed. The factor structure of PAS-TR determined by factor analysis was found to have construct validity.

CFA supported the 2-factor structure of the scale obtained as the result of EFA. The evaluation did not reveal good compliance regarding the X2/df ratio and the RMSEA value. It has been stated in the literature that χ2/df<3 indicates a perfect fit, and the acceptable RMSEA value is ≤8 (34,35). The confirmatory factor analysis for this structure revealed the results as χ2=279.891, df=128 (p<0.001), χ2/df=2.19, RMSEA=0.063, GFI=0.905, CFI=0.949, IFI=0.950 and SRMR=0.049. These results confirmed the suitability of the 2-factor structure.

Reliability

Reliability was assessed using Cronbach’s α internal consistency coefficient, item-total correlation, and rest-retest analysis. Cronbach’s α consistency coefficient was calculated as an indicator of inner consistency and homogeneity of the PAS-TR. The scale is classified as quite reliable between 0.60 and 0.79, and highly reliable between 0.80 and 1.00 (36). Cronbach’s alpha coefficient was determined as 0.92-0.93 in the validity and reliability study of the original scale (29). In this study, Cronbach’s α coefficient was used for determining the internal consistency of the PAS. The internal consistency level was calculated as 0.925 for the first factor, 0.824 for the second factor, and 0.906 for the entire scale. When the corrected item-total correlation levels were investigated, it was found that the items with the highest level were items 5 and 8 (0.772 and 0.733, respectively), whereas those with the lowest level were items 9 and 12 (0.313 and 0.370, respectively). The internal consistency of PAS-TR was high. PAS-TR is a reliable scale having internal consistency.

The items of the scale having low coefficients are considered not reliable enough. When the correlation coefficient of the item is under 30, it is suggested that insufficient reliability is present (37). The item-total correlation scores of PAS-TR were 0.313-0.772. The total score correlation coefficients of entire items in the scale were above 0.30. From this aspect, the total correlation values of PAS-TR were at the suitable reliability level.

When PAS-TR was done 77 students with an interval of three weeks for test-retest analysis, high-level positive statistically significant relationships were found to be present regarding the entire scale and all its sub-dimensions. Consistency of PAS-TR with time was determined. This result revealed that the internal consistency of the scale was high and that reliable results could be obtained in more than one application of the scale (31).


Conclusion

The results of this study in which we had measured the validity and reliability of the PAS revealed that the PAS adapts well to Turkish society.

Ethics

Ethics Committee Approval: The present cohort study was designed as a survey and was approved by the Clinical Research Ethics Committee of Adnan Menderes University Faculty of Dentistry (decision no: 98318678-020, date: 28.03.2018).

Informed Consent: Consent was obtained from the questionnaire participants.

Peer-review: Externally peer-reviewed.

Authorship Contributions

Concept: R.Ç.A., Design: R.Ç.A., Data Collection or Processing: A.K., M.Ş.A., Analysis or Interpretation: S.Ç.S., F.A., Literature Search: R.Ç.A., A.K., M.Ş.A., Writing: R.Ç.A.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.


Images

  1. Matuszek S. Animal-facilitated therapy in various patient populations: systematic literature review. Holist Nurs Pract 2010; 24: 187-203.
  2. Beetz A, Uvnäs-Moberg K, Julius H, Kotrschal K. Psychosocial and psychophysiological effects of human-animal interactions: the possible role of oxytocin. Front Psychol 2012; 3: 234.
  3. Marcus DA, Bernstein CD, Constantin JM, Kunkel FA, Breuer P, Hanlon RB. Impact of animal-assisted therapy for outpatients with fibromyalgia. Pain Med 2013; 14: 43-51.
  4. Bert F, Gualano MR, Camussi E, Pieve G, Voglino G, Siliquini R. Animal assisted intervention: A systematic review of benefits and risks. Eur J Integr Med 2016; 8: 695-706.
  5. Straede CM, Gates GR. Psychological health in a population of Australian cat owners. Anthrozoös 1993; 6: 30-42.
  6. Serpell J. Beneficial effects of pet ownership on some aspects of human health and behaviour. J R Soc Med 1991; 84: 717-20.
  7. Odendaal JS. Animal-assisted therapy - magic or medicine? J Psychosom Res 2000; 49: 275-80.
  8. Parello CAK, Wesley Y, Campbell M. Examining pet attitude in relationship to loneliness and parenthood motivation in pet-owning adults. Health 2014; 6: 598-606.
  9. Jacobson KC, Chang L. Associations Between Pet Ownership and Attitudes Toward Pets With Youth Socioemotional Outcomes. Front Psychol 2018; 9: 2304.
  10. Amiot CE, Bastian B. Toward a psychology of human-animal relations. Psychol Bull 2015; 141: 6-47.
  11. Krause-Parello CA, Kolassa J. Pet Therapy: Enhancing Social and Cardiovascular Wellness in Community Dwelling Older Adults. J Community Health Nurs 2016; 33: 1-10.
  12. Min M, Zaw CC. A review on animal-assısted therapy and actıvıtıes for healthcare and teaching of children. Journal of Education and Social Scinces 2016; 5: 40-6.
  13. Guthrie MF, PH Marshall, Hendrick SS, Hendrick C, Logue E. Human love styles and attitudes toward pets. Anthrozoös 2018; 31: 41-60.
  14. Özen A, Özen R, Yaşar A, Armutak A, Bayrak S, Gezman A, Şeker İ. Attitudes of Turkish veterinary students and educators towards the moral status of animals and species rating. Kafkas Üniv Vet Fak Derg 2009; 15: 111-8.
  15. Doğan Ö, Özkul T, Özen A. The attitudes and views of Firat and Uludag University Faculty of Veterinary Medicine students on the subject of dog psychology. Ankara Üniv Vet Fak Derg 2011; 58: 85-91.
  16. Sıddıq AB, Erdem Ç, Şanlı Ş. The Scopes of Anthrozoology, the Study of Human-Animal Relationships in Turkey. İnsan ve Toplum Bilimleri Araştırmaları Dergisi 2018; 7: 805-26.
  17. Wilson CC, Netting FE, New JC. The pet attitude inventory. Anthrozoös 1987; 1: 76-84.
  18. Johnsons TP, Garrity TF, Stallones L. Psychometric evaluation of the Lexington attachment to pets scale (LAPS). Anthrozoös 1992; 5: 160-75.
  19. Staats S, Miller D, Carnot MJ, Rada K, Turnes J. The Miller-Rada commitment to pets scale. Anthrozoös 1996; 9: 88-94.
  20. Lago D, Kafer R, Delenay M, Connell C. Assessment of favorable attitudes toward pets: Development and preliminary validation of self-report pet relationship scales. Anthrozoös 1988; 1: 240-54.
  21. Wilson CC, Netting FE. The status of instrument development in the human-animal interaction field. Anthrozoös 2012; 25: 11-55.
  22. Fiocco AJ, Hunse AM. The Buffer Effect of Therapy Dog Exposure on Stress Reactivity in Undergraduate Students. Int J Environ Res Public Health 2017; 14: 707.
  23. Dixon-Preylo B, Arikawa H. Comparison of vegetarians and non-vegetarians on pet attitude and empathy. Anthrozoös 2008; 21: 387-95.
  24. Williams J, Lawrence A, Muldoon J. Children and their pets: exploring the relationships between pet ownership, pet attitudes, attachments to pets, and emphaty. Education and Health 2009; 28: 12-5.
  25. Al- Fayez G, Awadalla A, Templer DI, Arikawa H. Companion animal attitude and its family pattern in Kuwait. Society and Animals 2003; 11: 17-28.
  26. Karaçam Z. Adaptation of Scales to Turkish. Ebelik ve Sağlık Bilimleri Dergisi. 2019; 2: 28-37.
  27. Römpke AK. Adaptation of the Pet Attitude Scale to German. Society& Animals 2019; 27: 11-24.
  28. ITC. Guidelines for Translating and Adapting Tests (Second Edition). Indernational Journal of Testing 2018; 18: 101-34. Doi: 10.1080/15305058.2017.1398166
  29. Templer DI, Arikawa H. The pet attitude scale. In Blazina C, Boyraz G, Shen-Miller D (Eds.)," in The psychology of the human-animal bond: A resource for clinicians and researchers. New York: NY, Springer; 2011.p. 335-59.
  30. Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychological Assessment 1994; 6: 284-90.
  31. Şencan H. Sosyal ve Davranışsal Ölçümlerde Güvenilirlik ve Geçerlilik. Ankara: Seçkin Yayıncılık; 2005.
  32. Burns N, Grove SK. The practice of nursing research appraisal, synthesis, and generation of evidence. St Louis: MO: Saunders 2009.
  33. Streiner DL, Kottner J. Recommendations for reporting the results of studies of instrument and scale development and testing. J Adv Nurs 2014; 70: 1970-9.
  34. Klem L. Reading and understanding MORE multivariate statistics. Structural equation modeling, Grimm LG, Yarnold PR, Eds., Washington D. C.: American Psychological Association 2000.
  35. Alpar R. Çok Değişkenli İstatistiksel Yöntemler (3. Baskı). Ankara: Detay Yayıncılık; 2011.
  36. Polit DF, Beck CT. Essentials of nursing research: appraising evidence for nursing practice (7th ed.). Philadelphia, PA: Lippincott Williams & Wilkins; 2010.
  37. Akgül A. Tıbbi Araştırmalarda İstatiksel Analiz Teknikleri-SPSS Uygulamaları, Ankara: Emek Ofset Ltd. Şti.; 2005.