The study population were German speaking households with underage children. Data were collected by the Berlin-based market research company INFO Marktforschungsinstitut using the survey software keyingress (Ingress GmbH). The survey was run between August 3rd and 11th 2020. We used a mixed mode design that combined computer-assisted telephone interviews (CATI; n = 402) with a computer-assisted web survey (CAWI; n = 622). For the recruitment of the CATI subsample, a dual-frame-design was used to include households with landline and parents with mobile phone numbers. Dual-frame design weighting accounted for different selection chances of parents available by landline vs. mobile phones and for the specific characteristics of landline users and mobile phone users. The telephone surveys were conducted by trained and supervised staff. Participants in the web survey were recruited from an active online-access panel. Participants received incentives from the panel provider to compensate their participation in the survey according to a fixed scheme. Respondents with an unrealistically short completion time (n = 66) were excluded from the CAWI data set. To provide representative data, the total sample was recruited according to current micro-census quota for the German population for age, sex, household size, educational level and residency.
To increase representativeness and to limit bias, we applied post-stratification weighting for sociodemographic factors to account for disproportionalities between our recruited sample and the micro-census quota. The recruited sample was adjusted to the current micro-census quota in terms of parent age, parent gender, household size, parent educational level and residency. After an iterative weighting procedure, each case has received an individual weighting factor (rounded mean weight 1.000, rounded minimum weight 0.378, rounded maximum weight 3.648). In this paper, sample characteristics were presented both for the unweighted data and weighted data; statistical analyses were conducted with the weighted data set. The results and conclusion do not change when using the unweighted data.
Data were collected on parent age and sex, parental status (biological parent, step parent, other), marital status, number of children in the household, and children’s age and sex. Family socioeconomic status (SES) was classified as low, medium, or high on the Winkler Index according to German population-based reference data .
Parent-Related Risk Factors
In addition, parent-related risk factors for an increase in ACEs were assessed: the parent’s risk of alcohol abuse during the pandemic was assessed by the alcohol abuse module of the German version of the Patient Health Questionnaire (PHQ-D) [16, 17]; the presence of a mental disorder (“Have you ever been diagnosed with a mental illness by a physician?”), parent’s own history of child physical or sexual abuse (“In your childhood, have you been hit, punched or otherwise physically hurt during you childhood or adolescence, or have you been forced by someone to an unwanted sexual action during your childhood or adolescence?”), and the parent’s own experience of physical or sexual violence (“As an adult, have you been kicked, punched, or otherwise physically hurt, or have you been forced by someone to an unwanted sexual action?”) were assessed by self-report measures. Furthermore, parents were asked on the presence of a chronic or severe physical condition and whether they were among the risk group for COVID-19 [16, 18].
Pandemic-related experiences and stress
The Pandemic Stress Scale  was developed to assess COVID-19-related experiences and pandemic-related stress. COVID-19-related personal experiences were assessed in relation to the parents themselves or to other family or household members: By three items, contact to persons with a COVID-19-infection/hospital admission/death, was assessed. Furthermore, work-related data (short-time work, loss of job/work, severe financial loss) were collected. Furthermore, we asked the parents to indicate the month(s) with the subjectively highest burden by providing the months January 2020 till August 2020 separately, with the additional items “all months were equally stressful” and “no month was especially stressful”. For the assessment of pandemic-related stress, the parents rated the subjective burden of 13 restrictions (e.g. school closures) on a 5-point scale (anchors 1 = “not at all stressful”, 5 = “extremely stressful”). A higher sum score indicated a higher burden (Cronbach’s alpha = 0.94). Parents were asked to base their answers on the point at which they felt most stressed since the beginning of the pandemic. For a full presentation of this scale, please see Supplementary Material 1.
The Parental Stress Scale  is an 18-item self-report questionnaire on positive and negative perceptions of parenthood. Items are rated on a 5-point scale, with higher scores indicating higher parental stress. Parents were asked to rate their stress (1) at the time of the subjectively highest burden (Cronbach’s alpha = 0.88) and (2) in January 2020 (Cronbach’s alpha = 0.90).
The stress module of the PHQ-D  was used to assess general stress at the time of the subjectively highest burden. The module covers ten items on different psychosocial stressors, e.g., health concerns, concerns about weight or appearance, sexual problems or work-related stress (“How strongly did you feel impaired by the following problems?”). The items were answered on a three-point scale (0 = not impaired, 1 = a little impaired, 2 = strongly impaired). The sum score for general stress showed good internal consistency in our sample (Cronbach’s alpha = 0.81).
Parent mental health
The PHQ-4 , a four-item screening measure for generalized anxiety and depression, was used to measure parent mental health. Parents rated on a four-point scale how often they experienced symptoms (generalized anxiety: Cronbach’s alpha = 0.78, depression: alpha = 0.78, total score: alpha = 0.86) at the time of the subjectively highest burden.
We used a well-established single-item measure  to elicit parental self-ratings of health (“If you were to rate your general state of health on a scale from 0 to 10 (“0” meaning “couldn’t be worse” and “10” meaning “couldn’t be better”), how would you rate your current state of health?”). Subjective health was assessed (1) for the time of the highest burden and (2) for January 2020.
Adverse Childhood Experiences (ACEs)
To provide a comprehensive assessment of ACEs, we collected data on child abuse, neglect and household dysfunction [23, 24]. We adapted the items of the pediMACE [25,26,27]. Parents were asked to report the occurrence of ACEs for the children in their household. First, the occurrence of severe stressful life experiences (e.g. violence, abuse, neglect) was assessed, followed by ten items on specific subtypes of those events: five subtypes of child abuse (verbal emotional abuse towards the child, nonverbal emotional abuse towards the child, witnessing domestic violence, physical abuse, sexual abuse), three subtypes of neglect (emotional neglect, physical neglect, supervisory neglect), and two subtypes of household dysfunction (problems related to alcohol or substance use, mental illness in the household). Noteworthy, while the first item referred to severe forms of ACEs such as violence, abuse or neglect, item wordings of the subtypes mainly reflected low severity levels on the maltreatment classification system [23, 28]. Parents were first asked whether the distinct subtype of ACE had ever occurred in the child’s life and, if yes, to indicate the change in occurrence since the beginning of the pandemic on a five-point change scale (anchors “significantly more often” to “significantly less often”). For a full description of the items, see Supplementary Material 2.
Positive and negative experiences during the pandemic
In addition to the questionnaires, we asked two open questions on perceived highest burden and positive aspects of the pandemic: “Overall, what caused you the most stress during the pandemic?” and “What has changed for the better during the pandemic?”. Parents responded by typing their answers into a box (online survey) or by telling the interviewer (telephone survey).
The sample comprised 1024 parents with a mean age of 41.7 years (SD = 8.37; range 18–73; weighted M = 40.89, SD = 8.17, range 18–73). Mean child age was 9.41 years (SD = 4.78, range 0.5–17.0; weighted M = 9.19, SD = 4.78, range 0.5–17.0). Table 1 summarizes the sociodemographic characteristics of the study participants, including both the unweighted raw data and the weighted data. In addition, comparison data from the recent German micro-census  are included in Table 1.
Data on COVID-19-related experiences and parent-related risk factors are summarized in Table 2. Almost half of the sample identified April and May 2020 as the most stressful months. After that, a continuous decrease was observed until August 2020.
For data analysis and inferential statistics, the weighted data set was used. Outcome measures were analysed by descriptive statistics and compared with reference scores by t tests. Answers to the open questions were analysed by means of content analysis. Based on the first 100 answers, the team developed categories for negative and positive experiences in mutual discourse. Each answer could be classified to multiple categories. Two members of the team coded all answers independently. Discrepancies were discussed in the team until a final decision was reached. Frequencies of each category were calculated.