Survey data on the consequences of COVID-19 and home confinement on the educational community and families in Spain

Abstract

This paper presents a dataset concerning the consequences of the COVID-19 pandemic and home confinement on the educational community and families, and the possibilities and opportunities for the return to schools. Data were collected through an online based cross-sectional survey between June 29, 2020 and July 12, 2020 in Spain. A total of 7,305 people who had children in their care during the COVID-19 crisis and the home-confinement period responded to the survey. The survey contained items concerning (i) socio-demographic information, (ii) conciliation of work, personal and family life during confinement, (iii) the impact of the pandemic on the respondent's family, and (iv) the respondents' opinion on their child(ren)'s return to school. Data were analysed using Stata (version 14) and are represented as frequencies and percentages based on responses to the entire survey. Researchers can use the dataset to analyse how home confinement impacted people with children in their care. Additionally, government authorities and education policymakers can use the data to ensure that schools respond to parents' main concerns in a pandemic context, as well as to be prepared to implement appropriate protocols in possible future similar crisis.

Specifications Table

Subject	Education
Specific subject area	Consequences of the COVID-19 and home confinement on the educational community and families
Type of data	Primary data, tables
How data were acquired	Survey data were gathered using an online survey platform (google forms). The questionnaire is provided as a supplementary file
Data format	Raw. Analysed. Filtered (descriptive statistics)
Parameters for data collection	The survey data were obtained from 7,305 respondents living in Spain who had children in their care during the COVID-19 crisis and the home confinement period
Description of data collection	The data were obtained through an online questionnaire shared via e-mail, social networks (Instagram and Twitter) and WhatsApp
Data source location	Country: Spain
Data accessibility	Repository name: Mendeley Direct URL to data: https://data.mendeley.com/datasets/kbvg3j3h3k/2

Value of the Data

• •These data provide information on the consequences of the COVID-19 crisis and the home confinement period on the educational community and families, which is important for understanding the home confinement impact at a personal and family level. Additionally, the dataset provides information on parents' views on the return to school after the period of confinement.
• •During the COVID-19 crisis, especially at the beginning, many workers, both in private companies and in the public sector, were forced to telework from home. Teleworking had been an option demanded for years by the main unions and had been presented in several electoral programs. However, the nature of the COVID-19 crisis meant that, during the lockdown in Spain from March to June, schools were closed and parents had to deal with parenting and working at home at the same time. The data from this study show how difficult this situation was and how parents with young children especially suffered the consequences of house confinement during this period. Researchers can use the dataset to analyse how home confinement impacted people with children in their care. Additionally, the authorities can benefit from these data to ensure that schools respond to parents' main concerns in a pandemic context, as well as to be prepared to implement appropriate protocols in possible future similar crisis.
• •Other researchers around the world can use these data to conduct cross-cultural comparisons, examining similarities and differences in the consequences of home confinement on families with children across the world. Of course, in order to be able to carry out these analyses, it would be necessary to undertake a joint analysis with qualitative information in order to contextualise the data appropriately and make relevant comparisons.
• •The dataset enables subgroups comparison based on sociodemographic characteristics (e.g., gender, place of living, work situation, schooling stage of the child or children).

1. Data Description

The period of home confinement experienced in most countries in the first half of 2020 as a consequence of the COVID-19 health crisis has had psychological consequences for a large part of the population [1,2]. Moreover, expectations about their future have also been analysed in [3,4]. In this sense, children and adolescents have been one of the most affected population groups, as the closure of schools significantly altered their social and educational life [5]. Additionally, previous studies suggested that the confinement has had a great impact on the health-related behaviours of children [6,7]. Likewise, parents have been struggling to combine their jobs with the care of their children, as has been analysed for Canada in [8].

In view of this, this dataset provides relevant information on the consequences of the COVID-19 home confinement, ordered by the Spanish government between the 15th of March and the 21st of June, on the educational community and families and on the possibilities and opportunities for the return to schools. The survey involved 7,305 respondents living in Spain who had children in their care during the COVID-19 crisis and the home confinement period.1 The questionnaire and variables codebook are provided as a supplementary file.

The data include four major groups of variables. A first group of variables (A) refers to 16 items related to individual and family sociodemographic characteristics, including information on the gender of the respondent, place of current residence, characteristics of the living unit, work situation, schooling stage of the child or children, school ownership, and special educational needs. Table 1 shows the distribution of responses for all variables included in group (A).

Table 1.

Distribution of responses in relation to socio-demographic variables (A).

Variable	Freq (n)	% / Mean
Gender
Male	600	8.23%
Female	6,686	91.77%
Autonomous Community of residence
Andalucía	535	7.32%
Aragón	159	2.18%
Asturias	124	1.70%
Canarias	150	2.05%
Cantabria	67	0.92%
Castilla la Mancha	223	3.05%
Castilla y León	234	3.20%
Cataluña	715	9.79%
Ceuta	6	0.08%
Comunidad Valenciana	3,125	42.78%
Extremadura	57	0.78%
Galicia	212	2.90%
Islas Baleares	86	1.18%
La Rioja	31	0.42%
Madrid	1,173	16.06%
Melilla	4	0.05%
Murcia	104	1.42%
Navarra	64	0.88%
País Vasco	235	3.22%
Kind of place of current residence
Rural	1,027	14.45%
Small Town	1,805	25.40%
Big City	4,273	60.14%
Local Income	3,094	25,961.87
Ownership of child/children's educational establishment
Publicly-funded private	1,810	24.78%
Private	937	12.83%
Public	4,558	62.40%
Living unit during confinement
One adult person with a minor or minors in care	699	9.57%
Two adults with a minor or minors in their care	6,262	85.72%
More than two adults with a minor or minors in their care	344	4.71%
Family in charge of dependent persons
No	6,829	93.48%
Yes	476	6.52%
Respondent worked during confinement
No	2,526	34.58%
Yes	4,779	65.42%
Worked during confinement (other adult in the family)
No	1,292	18.69%
Yes	5,621	81.31%
Child/Children in first cycle of Early Childhood Education (0 to 2 years)
No	4,415	60.44%
Yes	2,89	39.56%
Child/Children in second cycle of Early Childhood Education (3 to 5 years)
No	3,713	50.83%
Yes	3,592	49.17%
Child/Children in Primary Education (6 to 12 years)
No	4,368	59.79%
Yes	2,937	40.21%
Child/Children in Secondary Education (12 to 16 years)
No	6,628	90.73%
Yes	677	9.27%
Child/Children in Baccalaureate (16 to 18 years)
No	7,194	98.48%
Yes	111	1.52%
Child/Children in Vocational Education
No	7,276	99.60%
Yes	29	0.40%
Child/Children with special educational needs
No	6,878	94.15%
Yes	427	5.85%

A second group of variables (B) refers to 19 items that measured the conciliation during the home confinement period including information on paid workload, housework, time spent helping children with homework, and time available for other activities such as sports or talking to friends. Table 2 shows the distribution of responses for all variables included in group (B).

Table 2.

Distribution of responses in relation to work conciliation during the confinement (B).

Variable	Freq (n)	%
Workload changed
No, I have worked the same hours	1,435	19.90%
Yes, I have lost my job	882	12.23%
Yes, I have worked more hours	2,307	32.00%
Yes, I worked less than usual	1,079	14.97%
Yes, I have voluntarily asked for a reduction	254	3.52%
Yes, I have voluntarily resigned from my job	217	3.01%
I work solely to care my family	1,036	14.37%
More housework and care work
No	586	8.03%
Maybe	440	6.03%
Yes	6,269	85.94%
I have slept
I have not been able to	155	2.14%
Less than before	3,917	54.01%
As before	2,240	30.88%
More than before	941	12.97%
I had leisure time
I have not been able to	2,509	34.55%
Less than before	3,231	44.50%
As before	615	8.47%
More than before	906	12.48%
I played sports
I have not been able to	2,934	40.40%
Less than before	2,185	30.08%
As before	1,140	15.70%
More than before	1,004	13.82%
I talked to my friends
I have not been able to	422	5.81%
Less than before	3,407	46.88%
As before	1,960	26.97%
More than before	1,478	20.34%
I have been in touch with my extended family
I have not been able to	699	9.61%
Less than before	2,332	32.05%
As before	2,479	34.07%
More than before	1,766	24.27%
I remembered things from the past
I have not been able to	517	7.13%
Less than before	485	6.69%
As before	2,747	37.87%
More than before	3,505	48.32%
I made decisions about the future
I have not been able to	1,066	14.71%
Less than before	852	11.75%
As before	2,699	37.24%
More than before	2,631	36.30%
I had sex
I have not been able to	1,248	17.30%
Less than before	2,294	31.80%
As before	2,952	40.93%
More than before	719	9.97%
Problems reconciling
No	1,526	21.11%
Maybe	861	11.91%
Yes	4,843	66.98%
Domestic and care help
No	5,850	80.24%
Yes	111	1.52%
I prefer not to answer	1,330	18.24%
Started day tired
No	1,209	16.55%
Maybe	668	9.14%
Yes	5,428	74.31%
Time off
No	4,109	57.26%
Occasionally	1,731	24.12%
Yes	1,336	18.62%
Interrupted working day to take care of children
No	2,055	29.09%
Yes	5,009	70.91%
If yes to interrupted, how often
Occasionally	835	16.21%
Several times during the working day	2,269	44.22%
Several times an hour	2,027	39.50%
Hours accompanying children in schoolwork
I do not have time for it	591	8.35%
1-2 hours per day	3,895	55.01%
3-5 hours per day	1,670	23.59%
All day	924	13.05%
Shared electronic devices with children
No	2,408	33.23%
Yes	4,838	66.77%
Delayed bedtime or brought forward wake-up time
No	2,448	34.13%
Yes	4,725	65.87%

Thirdly, a group of 34 variables (C) measured the consequences of the pandemic at a personal and family level for the respondent, paying special attention to how the pandemic had affected the child or children in their care. Fig. 1, Fig. 2, Fig. 3 show the distribution of responses for all variables included in group (C).

Fig. 1.

Distribution of responses in relation to the consequences of the pandemic for the family and the children: positive impact (C).

Fig. 2.

Distribution of responses in relation to the consequences of the pandemic for the family and the children: negative impact (C).

Fig. 3.

Distribution of responses in relation to the consequences of the pandemic for the family and the children: child/children behaviour (C).

Finally, (D) 43 items measured aspects directly related to children's education and the return to school. Respondents were asked, for example, what they thought their children missed the most and what main challenges they identified for the return to school in September. Table 3 and Fig. 5, Fig. 6, Fig. 7 show the distribution of responses for all variables included in group (D).

Fig. 4.

Distribution of responses in relation to education and return to school: child/children missed (D).

Table 3.

Distribution of responses in relation to education and return to school (D).

Variable	Freq (n)	%
Ability to help child/children with online education
None	205	2.92%
A little	1,363	19.39%
Enough	3,288	46.77%
A lot	2,174	30.92%
Exchanged words or met with child/children's teachers
Never	956	13.26%
Occasionally	3,984	55.27%
Weekly	1,814	25.17%
Daily	454	6.30%
Involved in parents groups in child/children's class
I have not had time for it	684	9.55%
Less than before	1,102	15.38%
As Before	3,880	54.16%
More than before	1,497	20.90%
Talked to other parents about return to school in September
Never	1,381	19.02%
Occasionally	4,554	62.71%
Weekly	1,049	14.45%
Daily	278	3.83%
If school fees, alternatives to avoid paying
No	1,201	22.12%
Yes	4,229	77.88%
Collaboration of families: important role in return to school
No	146	2.01%
I do not know	1,076	14.79%
Yes	6,051	83.20%
Complement teaching with other activities outside the school
No	769	10.69%
Maybe	2,127	29.52%
Yes	4,309	59.81%

Fig. 5.

Distribution of responses in relation to education and return to school: challenge in school (D).

Fig. 6.

Distribution of responses in relation to education and return to school: level of priority (D).

Fig. 7.

Distribution of responses in relation to education and return to school: school space (D).

2. Experimental Design, Materials and Methods

The survey was developed in the early stages of the COVID-19 pandemic and adopted a descriptive online cross-sectional survey design to assess the consequences of the COVID-19 home confinement on the educational community and families, and to explore parents' views on their children's return to school. A total of 7305 participants living in Spain who had children in their care during the COVID-19 crisis and the home-confinement period responded to the survey from June 29, 2020 to July 12, 2020.

In order to collect and manage the data, the following steps were followed: (1) definition of the research objectives; (2) design of the questionnaire; (3) questionnaire pilot testing (validity, reliability, repeatability); (4) dissemination of the questionnaire; (5) collection and organisation of the data; and (6) interpretation of the information obtained. The questionnaire designed consisted of closed-ended question types (multiple-choice, “yes” or “no” and skip logic) and was created in two languages (Spanish and Valencian) using google forms. The link generated was shared by email to schools, on social networks (Instagram and Twitter) and via WhatsApp. The collected data were exported to Excel spreadsheets and data were analysed using the Stata software.

As regards the structure of the questionnaire, it consisted of a total of 50 questions and five sections: (1) the first section presented general information for the respondent on the purpose and functioning of the survey; (2) the second section asked 9 questions related to socio-demographic information of the respondents; (3) section three presented 19 questions on the conciliation of work, personal and family life during confinement; (4) in section four, 5 questions asked about the impact of the pandemic on the respondent's family; and (5) finally, in Section 5, 17 questions asked about the respondents' opinion on their child(ren)'s return to school.

In the process of designing, conducting and analyzing the survey, checks were carried out on the validity and reliability of the survey data. The first requirement that was considered is that of content validity to ensure the adequacy to the research objectives, which was analysed together with the other two key determinants in the development of any research: the resources (material, economic and human) and the time available to carry it out. As is often the case, in order to assess this type of validity, expert judgement was used to carry out an assessment by people qualified in the subject [9]. The expert panel consisted of 10 experts: 2 sociologists, 4 school teachers, 2 experts in education and 2 experts in quantitative methodologies. The evaluation was carried out in three rounds in which each of them evaluated the survey, making proposals for modifications so that once they had been integrated into a new questionnaire, they were again submitted to the judges for evaluation. All of them were given the questionnaire and a document to evaluate the following aspects: coverage of the proposed objectives with the questions included in the questionnaire, detection of redundant items, appropriateness of the language, order of the questions, appropriateness of the scales and response time.

Along with this analysis of adequacy to the objectives, the external validity and internal consistency of the survey were taken into consideration [10]. With respect to external validity, that which affects the possible generalizability of the survey results, we worked on the representativeness of the sample, i.e., the extent to which the sample has been able to represent, on a small scale, the variety of units that make up the study population.

In order to justify external validity, it is first necessary to explain the data collection process. Data collection mechanisms had to be selected to strike a balance between the possibility of finding a sufficient number of responses in the designed strata and also a balance between them. Procedures based on exhaustive lists of potential participants were discarded, which, while allowing for good sampling control, were not feasible due to the impossibility of having such lists. The alternative of using “snowball” mechanisms using contacts in schools and social networks was envisaged as a way of obtaining a sufficiently large sample, although it was expected to be unbalanced between strata. This was indeed the case, obtaining a sample of 7,305 responses, but with a strong imbalance between strata. Although it is well known that the most efficient sampling design is to use a probabilistic procedure with random selection of respondents, for the reasons already described this was not possible in our study. Instead, we resorted to non-probabilistic techniques that provide good results in situations with sufficiently large sample sizes and adequate weighting to balance the final estimates. Therefore, it was very important to carefully choose the variables that would generate the strata in the study population. After a review of the literature and discussion with the same group of experts who participated in the revision of the questionnaire, it was determined that the non-observable variable that most conditioned the results was the socio-economic level of the families. The inclusion of direct questions in the questionnaire on this aspect, such as the level of studies, type of work, salary, etc., presented obvious difficulties linked to the response rate in these questions and the reliability of the answers. Therefore, we looked for variables that could be included in the questionnaire that would indirectly reflect this issue. It was considered that the variables geographical location and school ownership could together provide overall information on socio-economic status.

Therefore, to correct for imbalances in the sample profiles after data collection, a weighting factor was generated and strata were considered based on the combination of the information regarding the geographical location and school ownership. For each of the strata, sample sizes were compared with population sizes. Weighting values were calculated by dividing the population proportion by the sample proportion for each stratum and an upper bound of 3 was set to avoid over-representation of minority groups. Each individual was assigned the weighting value corresponding to the strata to which he or she belonged. It is important to note that the descriptive results presented in this paper have been obtained without using the weighting factor.

Finally, as for the internal consistency of the questionnaire, the possibility of using Cronbach's Alpha coefficient was ruled out. Instead, comparative statistical analyses were carried out between descriptive summary values and association between variables obtained by breaking down the sample into 6 subsamples obtained randomly from the overall sample. The individuals in each sub-sample were randomly selected in such a way as to maintain the proportions of the two key control variables considered, which were the autonomous community and the ownership of the school. The results showed acceptable stability in the results obtained between the subsamples.

Ethics Statement

The authors declare that this data collection does not need ethical approval from appropriate institutional review boards or local ethics committees.

CRediT authorship contribution statement

Guillermo Palau-Salvador: Conceptualization, Investigation, Methodology, Data curation, Writing – review & editing. Kas Sempere: Methodology, Data curation. Nerea Gómez-Fernández: Data curation, Writing – original draft. Ana Belda-Marco: Conceptualization, Investigation, Methodology. Isabel González-Galindo: Conceptualization, Investigation, Methodology. Miriam Hoyo-Juliá: Conceptualization, Investigation, Methodology. Davinia Ros-Bonanad: Conceptualization, Investigation, Methodology. José-Miguel Carot Sierra: Conceptualization, Investigation, Methodology, Data curation, Validation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships which have or could be perceived to have influenced the work reported in this article.