Demand for Tutoring - Evidence From a Parallel Survey Experiment

We investigate how information about the consequences of one year less of schooling and fewer skills learned affects the demand for tutoring lessons in two parallel survey experiments among adolescents aged between 14 and 17 years and their parents. Respondents in a randomly selected treatment group will be informed about the loss in lifetime income that can be associated with fewer years of schooling and decreased skill development. After information provision, respondents in the treatment group are asked about participation in tutoring lessons (for their child or themselves depending on the sample), their willingness-to-pay for one tutoring lesson, and the willingness to acquire more information about tutoring providers in Germany. Respondents in the control group answer the same questions, but without receiving information about the consequences on income.

Respondents will answer our survey questions and participate in the survey experiment via an online platform. For the purpose of the experiment, respondents will be randomly assigned to one of two experimental groups (control group and treatment group). Randomization in the adolescent and parent sample will be independent from each other.

[Information provision]: Parents and adolescents in the treatment group will receive the following information:
“According to calculations of a study, lifetime income, i.e. a person's earned income calculated over his or her entire working life, is reduced by 10 percent on average if a person attends school for one year less and thus learns fewer skills at school.
For a person who earns 2,100 Euro (net) per month, which is roughly the average income in Germany, this would correspond to a lower income of around 210 Euro per month.”

The information is based on calculations from Piopiunik, Kugler and Woessmann (2017) and Hanushek, Schwerdt, Wiederhold and Woessmann (2015) as well as Statista on the average monthly income in Germany.

[Outcome questions]: After information provision, members of the treatment group (parents and adolescents) will be asked about (i) how the COVID-19 pandemic affects the future income of their children or themselves, respectively, (ii) intended participation in a tutoring program for their child or themselves, respectively, (iii) information acquisition about tutoring programs in Germany as well as (iv) their willingness-to-pay for one tutoring session.

Members of the control group answer the same outcome questions without receiving the information.

References:
- Piopiunik, M., Kugler, F., & Wößmann, L. (2017). Einkommenserträge von Bildungsabschlüssen im Lebensverlauf: Aktuelle Berechnungen für Deutschland. ifo Schnelldienst, 70(07), 19-30.
- Hanushek, E.A., G. Schwerdt, S. Wiederhold und L. Wößmann (2015). Returns to Skills around the World: Evidence from PIAAC, European Economic Review 73, 103–130.
- Statista 2022: Höhe des durchschnittlichen Nettolohns/Nettogehalts im Monat je Arbeitnehmer in Deutschland von 1991 bis 2021. https://de.statista.com/statistik/daten/studie/370558/umfrage/monatliche-nettoloehne-und-gehaelter-je-arbeitnehmer-in-deutschland/ [accessed: 2022-06-28]

2022-06-28

2022-08-01

Our primary outcomes of interest are respondents’ expectation about the future income of their children/themselves, intended registration and participation in a tutoring program, information acquisition about tutoring programs in Germany as well as willingness-to-pay for one tutoring session.

We elicit planned registration/participation in tutoring lessons as follows:
[outcome1 - parents]
Tutoring and remedial programs often aim to help children and adolescents to complete tasks for school outside of regular class time and help them catch up on learning. These programs can take place online or on-site.

How likely are you to enroll your child in a tutoring or remedial program in the next 6 months?

[outcome1 - adolescents]
Tutoring and remedial programs often aim to help children and adolescents to complete tasks for school outside of regular class time and help them catch up on learning. These programs can take place online or on-site.

How likely are you to enroll in a tutoring or remedial program in the next 6 months?

Answers to both questions can be provided on a slider, ranging from very unlikely (0%) to very likely (100%).

The outcome question on the willingness-to-pay for a tutoring session is worded as follows:
[outcome2 - parents]
“We will now present you with 7 scenarios. In each scenario you can choose between an amount of money for yourself or a tutoring lesson (90 minutes) for your child at the tutoring program Schuelerhilfe.
Schuelerhilfe is a nationwide tutoring provider in Germany where your child can receive one-on-one tutoring in subjects such as math, German, English or French. The tutoring can take place both online and on-site.

After the survey, we will select one person from all respondents for whom one of the scenarios will be randomly chosen and implemented. If one of your scenarios is selected, either the money you would like to keep for yourself will be credited to your survey account after the survey, or you will receive an email with information about how your child can take advantage of the tutoring session.

Now please make a decision in each line.”

[outcome2 - adolescents]
“We will now present you with 7 scenarios. In each scenario you can choose between an amount of money for yourself or a tutoring lesson (90 minutes) for yourself at the tutoring program Schuelerhilfe.
Schuelerhilfe is a nationwide tutoring provider in Germany where your child can receive one-on-one tutoring in subjects such as math, German, English or French. The tutoring can take place both online and on-site.

After the survey, we will select one person from all respondents for whom one of the scenarios will be randomly chosen and implemented. If one of your scenarios is selected, either the money you would like to keep for yourself will be credited to the survey account after the survey, or we will send an email with information on how to claim the tutoring lesson.

Now please make a decision in each line.”

We then present seven scenarios with differing amounts of money, ranging from 0 to 30 Euro. For each scenario, respondents have to decide between the monetary reward and the tutoring lesson for their child or themselves, respectively.


In addition, we run a resurvey about one week after the main survey in which we re-elicit parents’ planned registration of their child for a tutoring lesson. The resurvey will allow us to investigate persistence of treatment effects of the first experiment.

To better understand the emergence of the treatment effects and to investigate potential mediating channels of our treatments, we ask a range of further outcome questions: (i) importance of different aspects for the respondents’ decision to participate in a tutoring program, (ii) information acquisition about tutoring programs in Germany (iii) expectation about the future income of their children/themselves (iv) the perceived impact of COVID-19 on students’ life.

In addition, we plan to perform heterogeneity analyses with respect to respondents’ prior beliefs as well as the extent to which the students’ learning has been affected by COVID-19.

We elicit respondents’ importance of different aspects for the their decision to register/participate in a tutoring program as follows:
[reasons - parents]
How important are the following considerations for your decision about whether to enroll your child in a tutoring program?
The consideration of whether
... my child has enough time to participate in a tutoring program.
... my child is motivated to participate in a tutoring program.
... my child would benefit from the tutoring program.
... we can financially afford that my child participates in a tutoring program.
... I support my child to participate in a tutoring program.
... my child needs help with schoolwork.

[reasons - adolescents]
How important are the following considerations for your decision about whether to enroll in a tutoring program?
The consideration of whether
... I have enough time to participate in a tutoring program.
... I am motivated to participate in a tutoring program.
... I would benefit from the tutoring program.
... my family can financially afford that I participate in a tutoring program.
... my parents support that I participate in a tutoring program.
... I need help with schoolwork.
Answers can be given on a 5-point scale, ranging from “very important” to “not important at all”.

We elicit information acquisition about different tutoring programs as follows:
[info acquisition – parents and adolescents]
Would you like to receive further information, e.g. on subjects offered, location and registration options of tutoring programs and learning apps or learning platforms in Germany?
o Yes
o No

If you indicate "yes", further information will be displayed at the end of the survey.

We also elicit posterior beliefs about the impact that the COVID-19 pandemic has on the future income of their child/themselves.

[future income - parents]
Many students learned fewer skills at school during the COVID-19 pandemic, for example, due to the loss of instructional time during school closures.
How do you think the change in schooling during the COVID-19 pandemic will affect your child's future income?

As a reminder: the average income in Germany is equivalent to about 2,100 Euro (net) per month.

I believe that my child's income …
o will decrease by ____ Euro (net) a month.
o will increase by ____ Euro (net) a month.
o will not change.

[future income - adolescents]
Many students learned fewer skills at school during the COVID-19 pandemic, for example, due to the loss of instructional time during school closures.
How do you think the change in schooling during the COVID-19 pandemic will affect your future income?

As a reminder: the average income in Germany is equivalent to about 2,100 Euro (net) per month.

I believe that my income…
o will decrease by ____ Euro (net) a month.
o will increase by ____ Euro (net) a month.
o will not change.

We also elicit respondents’ perceived impact of COVID-19 on students’ life as follows:
[aspects of life - parents]
Now it's time for your assessment of how COVID-19-related school closures will affect various aspects of your child's life since the pandemic began. To what extent do the following statements apply?
- My child will have worse opportunities on the job market due to school closures.
- My child will be less satisfied with his/her life in the long run because of school closures.
- My child will suffer from long-term health problems due to the school closures.

[aspects of life - adolescents]
Now it's time for your assessment of how COVID-19-related school closures will affect various aspects of your life since the pandemic began. To what extent do the following statements apply?
- I will have worse opportunities on the job market due to school closures.
- I will be less satisfied with my life in the long run because of school closures.
- I will suffer from long-term health problems due to the school closures.
Answer can be given on a 5-point scale from “strongly agree” to “strongly disagree”.

In addition, we elicit respondents’ prior beliefs about the loss in income that can be associated with fewer years of schooling and skill development as follows:
[Prior beliefs - parents and adolescents]
“Many studies look at how the skills learned in school affect later income on the labor market.
Imagine a person who earns 2,100 Euro (net) a month, which is roughly the average income in Germany.
By how many Euro do you think this person's income will change on average if he or she attends school for one year less and thus learns fewer skills at school?

I think that this person's income will decrease by ____ Euro.
(The answer "0" means that this person's income will not change).”

Finally, we also elicit respondents’ beliefs about the extent to which student learning has been affected by COVID-19. We do this by posing a series of questions.
First, we elicit how many weeks the child could not go to school due to the COVID-19 pandemic and the number of weeks in online schooling.
Second, we elicit whether the child has already participated in a tutoring program.
Third, we elicit respondents’ belief about the learning losses due to COVID-19 as follows:
[statement - parents]
How much does the following statement apply to your child?
"My child has learned much less than usual in school because of the COVID-19 pandemic."

[statement - adolescents]
How much does the following statement apply to you?
"I learned much less during the school closures than I usually do in school".
Answers could be given a 5-point scale: “fully applies”, “rather applies”, “rather does not apply”, “does not apply at all”, “neither nor”.
In our resurvey about one week after the main survey, we re-elicit parents’ beliefs about the loss in income that can be associated with fewer years of schooling and skill development.
In the resurvey, we additionally elicit (i) whether parents informed themselves about tutoring programs and (ii) the reasons for (not) doing so (open-ended question).

We conduct the survey experiment in a sample of 2,000 parents and their adolescent children between the age of 14 and 17 years. The survey is conducted in cooperation with a German survey institute, Bilendi & respondi. The recruitment of the parents (and their children) is managed by Bilendi & respondi, which collects the data via an online platform. That is, our participants answer the survey questions autonomously on their own digital devices. Randomization is carried out by Bilendi & respondi at the individual level, using a computer.

Our experiment is structured as follows:
Respondents (parents) will be randomly assigned to the treatment group (p=0.5) or the control group (p=0.5). Independently, adolescents will also be randomly assigned to the treatment group (p=0.5) or the control group (p=0.5).

Sequence of events in the treatment group:
1. Elicitation of weeks the child could not go to school due to the COVID-19 pandemic and the number of weeks in online schooling
2. Elicitation of beliefs about the learning losses due to COVID-19
3. Elicitation of participation in previous tutoring programs
4. Elicitation of prior beliefs about the loss in income that can be associated with fewer years of schooling and skill development
5. Information provision
6. Elicitation of posterior belief about the impact that the COVID-19 pandemic has on the child’s future income
7. Elicitation of planned registration/participation in tutoring lessons
8. Elicitation of willingness-to-pay for tutoring session
9. Elicitation of information acquisition about different tutoring programs
10. Elicitation of importance of different aspects for the decision to participate in a tutoring program
11. Elicitation of the perceived impact of COVID-19 on students’ life

Sequence of events in the control group:
1. Elicitation of estimate of school closures and online schooling during COVID-19 pandemic
2. Elicitation of assessment about learning less at school during COVID-19
3. Elicitation of participation in tutoring program
4. Belief elicitation about change in income due to one year less of schooling and the resulting fewer skills learned
5. Belief elicitation about change in own (child’s) income
6. Elicitation of probability to participate in tutoring program
7. Elicitation of willingness-to-pay for tutoring session
8. Elicitation of information acquisition
9. Elicitation of reasons to (not) participate
10. Elicitation of impact of COVID-19 related school closures on life aspects

Randomization is carried out by the survey company Bilendi & respondi, using a computer.

at the individual level

No

2,000 parents and 2,000 adolescents

2,000 in the parent sample and 2,000 in the adolescent sample

2,000 parents and 2,000 adolescent children in the age between 14 and 17 years, 1,000 of each are in the treatment group, 1,000 in the control group; randomization in the two samples is carried out independently

None