Measuring Non-Parental Care in the NLSCY: Content and Process Issues - August 1999

5. Review of survey methods: processes used to interview care providers

Previous

Contents

Next

The NLSCY must collect information from all types of non-parental care providers: those working within the child's home, those working in both regulated and unregulated family-care settings, and those working in day care centres. The provider may be either a relative or a non-relative. This requires a strategy for identifying them. The NLSCY would not randomly select care providers to survey, but would need to interview care providers of NLSCY children. This would allow for the linkage of non-parental care on child each outcome.

Several non-parental care provider surveys have previously been conducted in Canada (HRIB sector studies reports (1998) and Caring for a Living (1992)) and their experience in identifying respondents could help inform decisions on how to best approach contacting and interviewing providers. These surveys have interviewed both regulated and unregulated home providers and providers working in day care centres.

The NLSCY would potentially conduct a brief telephone survey of care providers of the NLSCY children. The method for obtaining a 10 minute telephone interview with the care providers of the NLSCY children must allow for interviews in all of the care scenarios previously described.

5.1 Interviewing regulated family-care providers

HRIB of HRDC conducted a sector study including a survey of regulated family providers in Canada. A list of 15, 939 providers was obtained and used as the sample frame. The sample for the survey consisted of approximately every 14^th licensed home child care provider in Canada. The overall response rate was 50 percent with a low of 0 percent (3 providers) from P.E.I. to a high of 83 percent in Nova Scotia. Overall, a total of 1,107 providers responded from a stratified, random sample of 2,209. The information was deemed reliable at a national level. Some provincial/jurisdictional information was also possible.

The strength of this survey method included the fact that the sample was derived from actual lists of already registered/regulated child care providers. As well, mail out surveys were used. This method is a cost-effective way to conduct a survey, although potentially low response rates can be a concern. Reminders about the survey were mailed out to providers two weeks after the initial survey was mailed out.

Because the questionnaire was a paper and pencil self-complete questionnaire, some potential biases among those not responding to the survey can exist. These include lower response rates from those who do not speak English or French as a first language or who have low literacy levels. As well, providers who were very committed to their job may have been more likely to fill out the questionnaire.

5.2 Interviewing unregulated family-care providers

HRIB hired Statistics Canada to conduct a pilot test to determine the best method to sample unregulated care providers. The approaches relied on parents in the Labour Force Survey (LFS) and used two methods to identify care providers:

Method 1. Parents in the LFS sample with children aged 0-12 were asked if they used unregulated non-parental care. If parents responded in the affirmative, they were asked to provide the interviewer with the name and phone number of the care provider. This method lead to a final response rate of only 36.2 percent. The expected response rate was 65 percent.

Method 2. LFS respondents were asked if anyone in their household provided unregulated non-parental care. This method lead to 1.3 percent of households responding that an unregulated care provider lived in the household. The expected percentage of households with an unregulated care provider was 3 percent.

The high costs of conducting this survey using the above two methods tested, combined with the low response rates, lead HRIB to examine a new methodology for identifying unregulated care providers. The new methodology used a random-digit dialing technique to first contact households, and then to ask if anyone in the household provided paid child care. The respondent was screened out if she provided regulated care. This methods was deemed as the most advantageous and was in the end the method that was used to survey unregulated care providers. A short telephone survey of unregulated providers was conducted and had a response rate of 70 percent. In that survey, the random digit dialing method worked well, although this issue is not applicable to the NLSCY because care provider information will be sought from the parents of the NLSCY children. However, the challenges faced in trying to contact unregulated providers does highlight some important issues relevant to the NLSCY.

In the past, there have been concerns about successfully collecting information from unregulated caregivers. Parents using unregulated care and caregivers providing unregulated care are often wary of being asked questions and response rates in these surveys have often been lower than expected (Child care provider methodology report; sector studies (HRDC), 1998)). Fears about being surveyed could arise for two reasons: (1) the parents and/or caregiver may be concerned about the provision of accurate financial receipts to parents for income tax purposes; and (2) the parents and/or caregiver may be concerned about their compliance with provincial regulations regarding adult-to-child ratios or other regulations in the care setting.

5.3 Interviewing centre-based providers

The 1992 survey of centre-based providers "Caring for a Living" conducted surveys with both centre directors and staff. Centres were sampled based on the number of care centres in each province, the auspice of the centres (profit or non-profit status of the centre), and the average number of workers percentre in each province or territory. Directors of centres were interviewed on the telephone and sent a short questionnaire. Centre staff were sent a questionnaire in the mail. Reminder letters and phone calls were used as a follow up for questionnaire completion.

A total of 969 centres were selected for the sample with an expected response rate of 50 percent. Directors were telephoned to see if they would participate in the survey. Participating directors and staff were then sent paper and pencil questionnaires. The response rate for the director's questionnaire was 52 percent with some variation across provinces (ranging from a low of 33 percent in the Northwest Territories to a high of 71 percent in Saskatchewan). Response rates for the teaching staff questionnaire ranged from a low of 21 percent in Alberta to a high of 42 percent in Saskatchewan, with an average response rate of 34 percent.

There were several advantages to using a combined method of telephone and mail out surveys. The telephone call could be used to establish the appropriate respondent and could serve to establish a rapport between the interviewers and the respondents. The mail out surveys provided a cost-effective method of obtaining information. The disadvantages to using mail out surveys were discussed in the previous section.

5.4 Summary

Overall, the above experiences in interviewing non-parental care providers can help inform decisions for the NLSCY. The difficult issue of contacting the appropriate respondent is not an issue for the NLSCY in the same way it was for the previous surveys. Earlier surveys were concerned about contacting providers that provided a specific type of care only. For the NLSCY, the interest is in interviewing providers who provide all types of care for its sample of children. Therefore, the appropriate source for the contact information is very clear: the parents of the NLSCY children. The difficulty lies in obtaining permission from parents to contact their child's provider. Parents may have concerns such as: Will the NLSCY be evaluating their child's provider as good or bad? Will the NLSCY be asking about the cost of care, when this could be a sensitive matter (especially for parents who do not receive receipts for the care)? These issues will have to be addressed in order for the NLSCY care provider survey to be successful.

Some of the previous surveys of care providers had some difficulty obtaining good response rates. This is a challenge for the NLSCY survey because there will be two occasions for refusal to participate in the survey which could substantially diminish the response rates. First, the NLSCY parent could refuse to provide permission for Statistics Canada to contact the provider. Secondly, the care provider herself could refuse to participate in the survey. This procedure parallels the pilot survey of unregulated providers using method 1. As mentioned earlier, that survey had a response rate of only 36.2 percent. However, the NLSCY is a longitudinal survey, so parents are familiar with the survey and its rationale. This may reduce the number of refusals. Several options for data collection are possible, but the NLSCY must choose a course of action that is cost-effective and complimentary to the main survey.

Earlier surveys used a mail out survey and some used both a mail out survey and a telephone interview. Non-parental care providers could also be surveyed using a face-to-face interview. Based on previous work, both the telephone and mail out survey methods were effective. The face-to-face method is always the preferred method, but it is also the most costly and therefore not always possible.