Psychometric Assessment of the Grandchildren Received Affection Scale

doi:10.21203/rs.3.rs-7525955/v1

This study aims to adapt the Grandchildren Received Affection Scale (GRAS). It is methodological research. Validity and reliability methods were used to analyze the data of this study on university students. Language and face validity for validity were determined as appropriate. Expert opinions were evaluated as similar for content validity. The reliability coefficient was 0.94. With confirmatory factor analysis, item coefficients were above 0.30. In addition, model fit indices were evaluated at appropriate values, and it was determined that all items and sub-dimensions of the Grandchildren Received Affection Scale were preserved. With the same analysis, the variance and eigenvalues of the GRAS were found to be within the appropriate limits. As a result, it can be said that GRAS is a tool suitable for Turkish culture and can be used easily.

Grandchildren

love

elder

The family is the minor structural unit of society, in which individuals maintain their interpersonal relations and development by establishing warm ties. The family is also a social, economic, and psychological solidarity system for its members. Today, in our society, where nuclear family structure is more common than extended family structure, it is observed that the bond with grandparents is gradually decreasing (Öztürk & Hazer, 2016). However, becoming a grandparent and experiencing the love of grandchildren is a very special feeling. After mother, father, and siblings, the role of grandparents is very effective in the raising and personality development of the child (Tunca & Durmuş, 2019).

While grandparents express that they like to spend time and play with their grandchildren, they report that it is a happy experience to love their grandchildren and sometimes help them take care of them (Hazer, 2012). Cultural characteristics passed down from grandparents to grandchildren make it easier for grandchildren to recognize and adapt to their family structures and societies. In addition, grandparents' love for their grandchildren provides them with spiritual satisfaction and helps them hold on to life by making them feel functional and useful (Hazer, 2012). Grandparents, as middlemen in family conflicts and intergenerational conflicts, and as guides to their children, provide a kind of counseling to their grandchildren and children in regulating the relationships between them and even between their children and their spouses. In this way, they resolve crisis situations and ensure that the family remains strong and that family dynamics are maintained in a healthy way. Due to these roles, grandparents are an important intermediaries in ensuring family continuity, preventing conflicts and overcoming crisis moments healthily. (Canatan, 2008), transferring family values and traditions to future generations, and contributing to the care of grandchildren (Sweeting et al., 2005).

One element that makes grandparents' lives meaningful is their children and grandchildren. In Turkish society, having a grandchild, spending time with a grandchild, and helping with his care are essential for family elders. Especially in families where both parents work, children who have not reached pre-school education age or after school spend the time their parents are at work under the supervision of an older person. Therefore, it is observed that grandparents play an essential role in the development of the child (Hazer, 2012). Because parents entrust their children to their first-degree relatives, they go to work comfortably, and the grandparents are happy because they feel helpful again and get to spend time with their grandchildren. Studies have determined that grandparents feel essential and accepted because of the need for them to take care of their grandchildren when their adult children are involved in work life (Çiftçi, 2008; Altan & Tarhan, 2018). Studies in the literature show that both parties benefit from the relationship between grandparents and grandchildren (Altan & Tarhan, 2018; Çiftçi, 2008; Duflos et al., 2020; Yavaş Çelik & Çiğdem, 2023 ). Therefore, it is essential to examine love between grandparents and their grandchildren and to have a tool to examine this love bond. This study aimed to adapt the Grandchildren Received Affection Scale (GRAS), which was developed to measure the love status of grandchildren by their grandparents in Turkish culture.

Questions of Research

The following questions will be tested in the research.

1. Does the factor structure of the Turkish form of the GRAS show similarities with the four-factor structure of the original scale (Mansson 2013)?

2. Is there a positive correlation between GRAS scores and Family Climate Scale (FCS) scores?

3. Are the Grandchildren the GRAS Cronbach alpha coefficients at acceptable levels?

4. Do the scores obtained on the GRAS have invariance over time?

Type of Research

A methodological research was conducted to adapt the Grandchildren Receive Affection Scale (GRASS) to Turkish culture.

Participants

The universe of the study consisted of all university students (undergraduate: 6093, associate degree: 3121, graduate: 443), while the sample group consisted of students (n=151) who volunteered to participate in the research, who used social media, and who could be reached by a simple random sampling method. The sample size should be determined as at least 5 and 10 times the number of items on the scale used (Seçer, 2015). Therefore, since the number of items is 17, samples should be between 105 and 170. The Kaiser-Meyer-Olkin (KMO) sample adequacy technique was used to determine sample adequacy, and data were analyzed when the KMO value was approached to one. It is recommended to use the Barlett test to determine the adequacy of the sample size before factor analysis, KMO, and the suitability of the sample for factor analysis (Ercan & Kan, 2004;Seçer, 2015). The KMO value takes a value between 0 and 1, which is expected to be above 0.6011. The sample was found to be sufficient. It was determined that the KMO coefficient was 0.91, and the chi-square value (X2=2238,469, p=0.0001) related to the Barlett test was at the p<0.05 significance level.

Data Collection

A socio-demographic questionnaire, the GRAS, and the Family Climate Scale (FCS) were used to collect data. The questionnaire and the questions on the scales were prepared in the virtual environment and delivered to the university students via social media. At the beginning of the form delivered to the students, the details of the study were explained, and voluntarism was questioned. If the student wanted to participate in the study, he filled out the questions and scales after giving his consent. The student ended the participation with the submit button at the end of the filled out form.

Question form: It was formed from the socio-demographic data of the participants and the questions that are thought to affect the grandparent-grandchild relationship (from the mother's side, from the father's side, living together, the frequency of interviews, the number of grandchildren they have, their age, etc.).

Grandchildren Received Affection Scale (GRAS): The scale was developed by Manssons. The scale consists of 17 items. The lowest possible score is 17 and the highest possible score is 119, and there is no cut-off point for the scale. In the scale development study conducted by Manssons (2013), the total Cronbach Alpha coefficient of the scale was found to be 0.91 (Mansson, 2013). In our study, the total Cronbach Alpha coefficient of the scale was found to be 0.94. Bu ölçeğin geçerlilik güvenilirlik çalışması yapıldı. Ayrıca Manssons (2015) geliştirdiği “The Ideal Grandparent Scale” isimli ölçeğinde geçerlilik güvenilirliği Türk toplumunda sınandı (Manssons,2015; Yavaş Çelik & Çiğdem, 2023)

The Family Climate Scale (FCS): The scale was developed by Björnberg and Nicholson (2007). The scale consists of 48 items (Björnberg & Nicholson, 2007). The validity, reliability study and psychometric examination of the scale in Turkish were conducted by Gönül, Işık-Baş, and Acar (2018). It was explained that the scale in the Turkish form of FCS consists of a total of 34 items and is a 5-point Likert type: There is no cut-off score for the scale. The total Cronbach Alpha coefficient of the scale was 0.91 (Işık-Baş & Acar, 2018). In this study, the Cronbach Alpha coefficient for the total scale was found to be 0.91. This scale was used in the study to ensure parallel form reliability.

Evaluation of Data and Analysis

First, the sample size was determined in Exploratory Factor Analysis (EFA) using Kaiser-Meyer-Olkin (KMO) and Barlett tests. Then, the explanation rates of each item's variance in a common factor were evaluated. When the variance values of the items were above 0.10, the reliability and validity analyzes of the scale were started. Language adaptation (translation back-translation technique), face validity, content validity (Kendall's W test), and construct validity were used to evaluate the scale's validity. Cronbach Alpha, test-retest and parallel form applications were used to evaluate the scale's reliability.

Ethical Dimension of Research

Ethics committee approval was obtained from the clinical research ethics committee. The study was based on volunteering and informed consent was obtained from the individuals. This study was conducted in accordance with the principles of the Declaration of Helsinki.

I. Validity

1. Language Adaptation

Translation-back translation technique was used for language adaptation. The scale questions were translated from English to Turkish by two experts who knew both languages and cultures. Later, translation from Turkish to English was made by two different experts in their fields. The translations were compared and turned into appropriate sentences

2. Face Validity

In addition, in order to evaluate the suitability of the language and face validity of the scale, the comprehensibility of the items in the scale was tested by 10 Turkish students studying at a university abroad and having good English. 18 days after this initial assessment, students were asked to evaluate the GRAS first in Turkish and then in English. Between the two assessments (first assessment: 86.50±22.98; second assessment: 81.27±22.16), students' scale mean scores were close to each other and there was no statistically significant difference between their mean scores (p>0.05).

3. Content Validity

The opinions of the experts were evaluated using Kendall's W test (p>0.05). The experts rated the scale items from (1 point) much change is needed to (4 points) very appropriate. It was found that the average score given by the experts to the scale items varied between 3.81±0.60 and 4.00±0.00 points.

4. Construct Validity

Confirmatory Factor Analysis (CFA): When the values determining the scale's suitability for Turkish culture were examined, item loads of the scale were ranged between 0.72 and 0.94 (Table 1). The model fit indices of the scale are; RMSEA: 0.04, NFI: 0.93, NNFI: 0.94, IFI: 0.95, RFI: 0.92, CFI:0.95, GFI:0.85, AGFI:0.87, RMR:0.25 and X2/SD:2.11 (Table 2).

II. Reliability

1. Cronbach Alpha

In the study, the total Cronbach Alpha coefficient of the scale was found to be 0.94, and the Cronbach Alpha coefficients of the sub-dimensions of the scale were found to be 0.95 in the love and respect sub-dimension, 0.90 in the compassion sub-dimension, 0.88 in the memories and humor sub-dimension, and 0.79 in the celebration sub-dimension.

2. Test-Retest

The test-retest was performed in a separate sample group (n=55) for reliability. Students were asked to re-evaluate the scale 15 days after evaluating it. Between the two evaluations (first evaluation: 87.50±27.98 and second evaluation: 86.27±21.16), it was seen that there was no statistically significant result between the averages of the scores given by the students to the scale items, and their answers were close to each other. (p>0.05).

3. Parallel Form Reliability

Family Climate Scale was used for parallel form reliability. Pearson product-moment correlation analysis showed a strong positive correlation between the two scales (r=0.85, p=0.001).

The study aims to adapt the Grandchildren Received Affection Scale (Mansson, 2013) developed by Manssons (2015) to Turkish culture. The findings of this study are similar to research of Manssons (2013) who developed GRAS, and Duflos et al., who adapted GRAS to French culture. (2020), and all sub-dimensions and GRAS items were preserved as in these studies (Mansson, 2013; Duflos et al., 2020). In this respect, it can be said that GRAS is a reliable and valid tool. In the validity and reliability studies of the scale, it should be ensured that the diversity between the two cultures is minimized and that the translation is understandable (Öner, 2009). The translation-back translation technique is recommended as a widely used method in language adaptation (Aksayan & Gözüm, 2002;Öner, 2009). When we look at the analyzes made, the results obtained in language adaptation and face validity, which are the first methods used in scale adaptation, are pretty satisfactory. The language adaptation of GRAS, whose original language is English, was carried out using the back translation method. First of all, GRAS was translated from English to Turkish independently by two experts who have a good level of English and know both cultures. The researchers examined all translations and compared them with the original text. Later, it was independently back-translated by two experts in the field of health who knew the two languages well, and it was determined that there was no difference in meaning between the two translations. In addition, to evaluate the suitability of the language validity and the face validity of the scale, the comprehensibility of the items in the GRAS was tested by ten Turkish students who speak English and are studying abroad. The students were asked to evaluate the scale in Turkish first. The same student group was asked to answer the GRAS in English after an 18-day break. It was seen that there was no statistically significant difference between the two assessments, and their answers were close to each other.

According to these results, it can be said that the language of GRAS was implemented understandably and validly while adapting it to Turkish. Content validity and expert opinions were evaluated in the next stage, the validity analysis. Content validity is used to determine how much the scale as a whole and each item in the scale explains the subject to be evaluated (Ercan & Kan, 2004; Vehid & Eral, 2014). In this context, the opinions of the experts consulted for content validity were evaluated using Kendall's W test. As a result of this test, it was determined that the experts' opinions did not change from each other, that they agreed, and that there was no statistically significant relationship between the opinions, and it was concluded that the content validity of the GRAS was appropriate. In the next step, the KMO value of GRAS and its suitability for CFA were evaluated. As a result of this evaluation, appropriate sampling was reached, and CFA was applied. As a result of this examination, it can be said that the original structure of GRAS has been preserved.

In addition, the total variance of the TSAS, whether its sub-dimensions were preserved or not, and the suitability of the item loads were tested, and it was determined that the specified (52%, 75%) (Seçer, 2015) value was above the variance value and the item loads of the scale were above 0.30 (Seçer, 2015). A CFA with construct validity is performed to determine whether the data in the hands of researchers fits the original construct. The researcher tests whether the data he has is compatible with the factor structure that has been constructed before. CFA is a validation method used primarily in adapting measurement tools developed in other cultures and samples. In other words, CFA is an analysis to evaluate the extent to which the factors formed from many variables, supported by an institutional basis, are compatible with the actual data (Ercan & Kan, 2004). In our study, the model fit indices of the scale values were found to be at the desired values due to the CFA analysis. In this context, it can be said that the fit indices and item loads of the GRAS are within valid limits. Cronbach's Alpha was tested first for reliability. In the analysis, it was determined that the Cronbach's Alpha coefficient of the GRAS was within the desired limits and was accepted as reliable.

The Cronbach Alpha coefficient is frequently used in rating scales, that is, in Likert type scales. The compatibility between the items on the scale is determined by calculating the Cronbach Alpha coefficient (Çakmur, 2012; Seçer, 2015). The Cronbach Alpha coefficient should be 0.60 or above. It is evaluated as high between 0.80–1.00, highly reliable between 0.60–0.80, low between 0.40–0.60, and unreliable between 0.00–0.40 (Karagöz, 2017). After the Cronbach Alpha reliability analysis, the reliability of the GRAS was evaluated by measuring its invariance over time. In a separate sample group (n = 55), the test-retest application was performed for reliability. The mean of the GRAS between the two measurements was compared with a 15-day break, and no significant statistical value was found between the two averages. According to this result, it can be said that GRAS is invariant over time. The parallel form was used for the final reliability analysis. Parallel form reliability looks at the correlation between another measurement tool, which is known to measure the same or similar content, and the measurement tool used in the same group.

The obtained correlation value indicates the stability level of the measurement result (Seçer, 2015). In this study, FCS was used as a parallel form, and it was determined that there was a strong positive correlation between the two scales. So, the FCS, which evaluates a situation similar to ours, and the GRAS scale have become useful tools for figuring out how people from different generations get along with each other.

Although there is a strong bond between grandchildren and grandparents in our country, and grandparents have an essential place in their grandchildren's upbringing, there are not many developed tools to evaluate the relationships between these two. Therefore, GRAS is a crucial tool for evaluating this relationship. In this study, we concluded that GRAS is an appropriate scale used in Turkish culture and that all its items and sub-dimensions are preserved. In this respect, GRAS can be used safely in future research and clinical evaluations.

Author Contribution

Author Contributions: “Conceptualization, ZC,MYC..; methodology, ZC,MYC.; software, ZC,MYC.; validation, ZC, MYC. formal analysis, MYC.; investigation, ZC,MYC.; resources, ZC,MYC.; data curation, ZC,MYC.; writing—original draft preparation, ZC,MYC.; writing—review and editing, MYC.; visualization, MYC.; supervision, MYC.; project administration, MYC.; funding acquisition, ZC,MYC.

Acknowledgement

na

Financial Resource

During this study, no financial support was received from institutions or organizations

Aksayan, S., & Gözüm, S. (2002). Kültürlerarası Ölçek Uyarlaması için Rehber I. Ölçek Uyarlama Aşamaları ve Dil Aşamaları. Hemşirelikte Araştırma Geliştirme Dergisi, 4, 9–20.
Altan, S., & Tarhan, S. (2018). Büyükanneden Toruna: Yaşamboyu Öğrenme Sürecinde Değer ve Beceri Aktarımı. Bartın University Journal of Educational Research/Bartın Üniversitesi Eğitim Araştırmaları Dergisi, 2(1), 26–42.
Björnberg, Å., & ve Nicholson, N. (2007). The family climate scales: Development of a new measure for use in family business research. Family Business Review, 20, 229–246. 10.1111/j.1741-6248.2007.00098.x
Canatan, A. (2008). Toplumsal değerler ve yaşlılar. Yaşlı Sorunları Araştırma Dergisi, 1(1), 62–71. Erişim adresi:http://dergipark.gov.tr/yasad/issue/21799/234272
Çakmur, H. (2012). Araştırmalarda Ölçme-Güvenirlik-Geçerlilik. TAF Preventive Medicine Bulletin, 11(3), 339–344.
Çiftçi, A. (2008). Ailede yaşlı bireyler. V. Aile Şurası Aile Destek Hizmetleri Bildirileri. T. C. Başbakanlık Aile ve Sosyal Araştırmalar Genel Müdürlüğü. Ankara. http://ailetoplum.aile.gov.tr/data/54293ea2369dc32358ee2b25/kutuphane_52___v__aile_surasi____ aile_destek_hizmetleri____bildirileri.pdf.
Duflos, M., Giraudeau, C., & Ferrand, C. (2020). What is emotional closeness between grandparents and their adolescent grandchildren? A systematic review. Journal of Family Studies, 0(0), 1–23. https://doi.org/10.1080/13229400.2020.1752771
Ercan, İ., & Kan, İ. (2004). Ölçeklerde Güvenilirlik ve Geçerlilik. Uludağ Üniversitesi Tıp Fakültesi Dergisi, 30, 211–216.
Hazer, O. (2012). Büyük ebeveynlerinin ergenler tarafından algılanan özelliklerini etkileyen faktörlerin incelenmesi. Hacettepe Üniversitesi Edebiyat Fakültesi Dergisi, 29(1), 123–141. Erişim adresi:http://dergipark.gov.tr/huefd/issue/41214/511219
Işık-Baş, G., & Acar, Ş. (2018). Aile İklimi Ölçeği’nin Türkçeye Uyarlanması ve Psikometrik Açıdan İncelenmesi. Türk Psikolojik Danışma ve Rehberlik Dergisi, 8(50), 165–200.
Karagöz, Y. (2017). SPSS ve AMOS Uygulamalı Nicel-Nitel Karma Bilimsel Araştırma Yöntemleri ve Yayın Etiği (pp. 24–35). Nobel Yayın Dağıtım.
Mansson, D. H. (2013). The Grandchildren Received Affection Scale: Examining Affectual Solidarity Factors. Southern Communication Journal, 78(1), 70–90. 10.1080/1041794X.2012.729124
Mansson, D. H. (2015). Developing the Ideal Grandparent Scale: Initial validity and reliability tests. Journal of Intergenerational Relationships, 13(3), 207–226. https://doi.org/10.1080/15350770.2015.10 58212
Öner, N. (2009). Türkiye’de kullanılan psikolojik testler. 2nd ed. İstanbul: Boğaziçi Matbaası. pp. 520–5.
Öztürk, S., & Hazer, O. (2016). Youth Perspectives on Intergenerational Solidarity in Families/Genç Nesil Bakışıyla Ailelerde Nesiller Arası Dayanışma. Turkish Journal of Family Medicine and Primary Care, 10(4), 22–232. 10.21763/tjfmpc.271095
Seçer, İ. (2015). SPSS VE LISREL ile Pratik Veri Analizi. Ankara: Anı Yayıncılık, pp. 211–58.
Sweeting, H., Ross, N., Hill, M., & Cunningham-Burley, S. (2005). Grandparents and teen grandchildren: Exploring intergenerational relationships. ESRC, Centre for Research on Families and Relationships.
Tunca, A., & Durmuş, E. (2019). Büyükannelerin Torun Büyütme Yaşantılarının İncelenmesi. SEFAD, 41: 209–226 e-ISSN: 2458-908X DOI Number: https://doi.org/10.21497/sefad.586615
Vehid, H. E., & Eral, G. (2014). Nitel Verilerin Değerlendirilmesinde Uygulanan İstatistiksel Yöntemler. Çocuk Dergisi, 14, 60–61.
Yavas Celik, M., & Cigdem, Z. (2024). The Ideal Grandparent Scale: Validation in a Turkish Young Adults Population. Ageing International, 49(2), 407–416.

Table 1. Item loads and cronbach alpha values of GRAS

Sub-dimension	Item of GRAS	Item point	Cronbach Alpha
love and esteem	I1	0.94	0.94	0.95
	I2	0.93	0.93
	I3	0.91	0.94
	I4	0.93	0.93
	I5	0.91	0.93
Caring	I6	0.85	0.94	0.90
	I7	0.83	0.94
	I8	0.83	0.94
	I9	0.90	0.94
	I10	0.80	0.94
Memories and Humor	I11	0.72	0.94	0.88
	I12	0.83	0.94
	I13	0.93	0.94
	I14	0.92	0.94
Celebratory	I15	0.77	0.94	0.79
	I16	0.77	0.95
	I17	0.88	0.95

Table 2. Validity and reliability analyzes of GRAS

Analyzes made	Results
Validity Analysis
Language adaptation	A translation-back translation technique was used for language adaptation. The scale items were translated from English to Turkish by two experts who knew the English language and the cultural structure of the two communities well. Then, the translations made from Turkish to English by two different experts in their fields were compared, and appropriate sentences were arranged.
Face validity(n=10)	The comprehensibility of the items in the scale was tested by 10 Turkish students studying abroad. Students were first asked to evaluate GRAS in Turkish. They were asked to evaluate the GRAS in English 18 days after the initial assessment. Between the two evaluations (first evaluation: 86.50±22.98, second evaluation: 81.27±22.16), it was seen that there was no statistically significant difference between the average scores given by the students to the scale questions and their answers were close to each other (p>0.05).
Content Validity	It’s evaluated using Kendall's W test. As a result of this test, it was determined that the opinions of the experts did not change between each other, they were in agreement, and there was no statistically significant relationship between the opinions (p>0.05).
Construct validity (n=151)	Exploratory Factor Analysis (AFA) KMO=0.91, X²=2238,469, p=0.001, item loads of the TSAS were found to be between 0.65 and 0.88, explaining 78.31% of the total variance of the scale. Confirmatory Factor Analysis (DFA) It was found that the item loads of the GRAS ranged from 0.72 to 0.94 (Table 1). The model fit indices of the scale are; RMSEA: 0.04, NFI: 0.93, NNFI: 0.94, IFI: 0.95, RFI: 0.92, CFI:0.95, GFI:0.85, AGFI:0.87, RMR:0.25 and X2/SD:2.11
Reliability Analysis
Cronbach Alpha	The total Cronbach Alpha coefficient of the GRAS was found to be 0.94. The Cronbach Alpha coefficients of the sub-dimensions of the scale are; 0.95 in the sub-dimension of love and esteem, 0.90 in the sub-dimension of caring, 0.88 in the sub-dimension of memories and humor, and 0.79 in the sub-dimension of celebratory.
Test-retest (n=55)	A different group of students (n=55) were asked to evaluate the scale twice with an interval of 15 days. Between the two evaluations (first evaluation: 87.50±27.98 and second evaluation: 86.27±21.16), it was seen that there was no statistically significant result between the average scores given by the students to the scale items and their answers were close to each other (p>0.05).
Parallel form reliability	FCS was used for parallel form reliability. Pearson product-moment correlation analysis showed a strong positive correlation between the two scales (r=0.85, p=0.001).

No competing interests reported.

Psychometric Assessment of the Grandchildren Received Affection Scale

Status:

Version 1

Abstract

INTRODUCTION

METHODS

RESULTS

DISCUSSION

CONCLUSION

Declarations

Author Contribution

Acknowledgement

References

Tables

Additional Declarations

Status:

Version 1