Online English Conversation Lessons: From What Age? When They Work and When They Don't
Why this topic is on your mind
A question we often get from readers of the sister article How should you choose an English conversation school? is: “In-person schools cost ¥10,000 to ¥20,000 a month, but with online English conversation you can do it every day for a few thousand yen a month. What do you think?”
Online English conversation certainly looks like an attractive option.
- Much cheaper than in-person schools
- Done entirely at home, with no burden of taking your child there and back
- Easy to keep up a little every day
- Less affected by health or the weather
On the other hand, anxieties also come up at the same time: “Does it really work through a screen?” “Won’t it end up just being a matter of showing my child a smartphone or tablet?” In this article, we answer those anxieties from the side of research. Rather than the language of advertising, we set out “under what conditions they work, and what ways of using them are less effective”.
First, the basic principle: language learning needs “social interaction”
Before getting into online English conversation, let us confirm just one fundamental principle of language learning. Once you grasp this, everything that follows becomes clear.
An experiment exposing American infants aged 9 to 10 months, growing up in an English-speaking environment, to Chinese (Mandarin) over 4 weeks gives one answer to this.
In the study, the infants were divided into three groups.
- Group A: read picture books and played with an adult who spoke Chinese (with interaction with a live person)
- Group B: shown a recorded video of the same Chinese speaker
- Group C: heard only the audio of the same Chinese
After 4 weeks, when a test of discriminating sounds specific to Chinese (consonants not found in English) was carried out, only the infants in Group A had acquired the ability to discriminate the Chinese sounds. Groups B and C did not differ from a control group that had no exposure to Chinese at all.
Reading this far, you might think “so screens are no good after all”. But the story goes one step further.
Two-way video chat is “close to face-to-face”: the impact of the Roseberry study
Kuhl’s study was in 2003. More than ten years later, in an era when video chat such as Skype had become widespread, the following study was carried out.
A study comparing whether toddlers aged 24 to 30 months could learn new verbs in three groups: (A) a group interacting face to face, (B) a group interacting with the same adult over video chat (real-time, two-way), and (C) a group shown a recorded video of that adult showed results of decisive importance for thinking about today’s online English conversation lessons.
The results were as follows.
- (A) Face-to-face group: able to learn the new verbs
- (B) Video chat group (real-time, two-way): able to learn to the same degree as the face-to-face group
- (C) Recorded video group: almost no learning effect was seen
In other words, even “through a screen”, if there is someone who responds in real time to the child’s reactions, learning effects close to face-to-face interaction appear. The difference between “recorded video”, which had no effect in Kuhl’s study, and “two-way video chat”, which had an effect in Roseberry’s study, lies in “whether the other person responds in real time to the child’s reactions (socially contingent interaction)”.
Putting these two studies side by side makes the criteria for evaluating online English conversation clear. “Materials that play recorded videos” cannot be expected to do much in research terms. On the other hand, “real-time, two-way video lessons” can function well enough as a substitute for face-to-face lessons.
From what age can children start?
Now for the practical question. When thinking about “from what age?”, the fact that the participants in the Roseberry study were 24 to 30 months old (2 to 2 and a half years) is one point of reference. However, this was a study measuring “whether language learning effects appear”; it does not mean that “children can start commercial online English conversation lessons from age 2”.
As a practical guide, it can be summarized as follows.
- From around ages 3 to 4, some children can start
- However, the premise is whether they can stay seated in front of the screen and whether they can respond to the teacher’s questions
- Up to 15 to 25 minutes per session is realistic
- Keeping a child who doesn’t want to seated and continuing goes against the research findings (Pinter, 2017)
What Pinter’s ‘Teaching Young Language Learners’ emphasizes is that in young children’s second language learning, creating experiences in which “the child has fun and wants to see the teacher again” matters more than anything. This is the same whether in person or online. “Whether a conversation through the screen is an enjoyable experience for this child right now” is a more practical basis for judgment than “from what age is it possible”.
Relationship with the American Academy of Pediatrics (AAP) screen time guidelines
What is often referred to regarding “showing screens to children” is the ‘Media and Young Minds’ policy statement. The AAP states:
- Under 18 months: avoid use of digital media, except video chat
- 18 to 24 months: watching high-quality programs together with a parent is acceptable
- Ages 2 to 5: no more than 1 hour a day of high-quality programs, preferably with a parent
What deserves attention is that the AAP explicitly states that “video chat is an exception”. This shows that the American Academy of Pediatrics’ position, based on findings such as the Roseberry study, is that “two-way conversation should be treated as a separate category from mere screen viewing”.
In other words, the simple idea that “screens = bad” and “screen time = something to cut” has already been abandoned by research and guidelines too. On the other hand, there is still the guide of no more than 1 hour a day even at age 5. A design that keeps online English conversation to about 25 minutes per session, several times a week, is consistent with this guideline.
When they work and when they don’t
Based on the research findings and practical guides, the cases where online English conversation works and where it doesn’t can be clearly contrasted.
Ways of using online English conversation that are less effective
Less effective
The child is seated in front of the screen while the parent is in the next room. Sessions are long, 40-50 minutes, and the child's concentration has run out in the second half. The teacher talks one-sidedly, and the child has few chances to respond. The child just stares blankly at the screen and hardly speaks. A section of watching recorded videos takes up most of the lesson. Continuing 'because we promised to do it every day' even though the child doesn't want to.
Ways of using online English conversation that are more effective
More effective
Keep each session to 15-25 minutes or less. Choose a teacher who watches the child's reactions closely and responds. At first, a parent sits beside the child and helps with reactions. The child has many chances to respond out loud, and the teacher waits for the child to speak. After the lesson, the child can say 'that was fun'. Time playing recorded materials is supplementary, and conversation is central.
The key points are three: “two-way interaction”, “amount of conversation”, and “a length the child can stay focused for”. In light of Kuhl’s and Roseberry’s research, the criterion that makes sense in research terms is to look at “how much the child can take part in responsive exchanges” rather than “how long the child is shown a screen”.
Actually, we tried a trial lesson, but my child froze in front of the screen and could hardly answer anything. Is there any point in continuing?
It’s too early to judge from one trial. Everyone is nervous at first. In fact, for a 5-year-old, talking with an unfamiliar adult through a screen is a higher hurdle than adults imagine. For the first few sessions, if you sit beside your child and step in as an interpreter and supporter, saying things like “That’s pink, isn’t it?” or “Shall we try saying ‘Yes’?”, it makes a huge difference to how secure your child feels.
I also feel that having a parent beside them might get in the way of my child’s independent learning…
I understand that worry, but in research terms it’s actually the opposite. What Roseberry’s study showed is that “responsive exchanges are what work”. If a responsive environment, including the parent, is in place, it becomes the foundation for learning effects. Once your child gets used to it, the parent naturally fades out: that order is fine.
The reality of “Cebu lessons” and “Filipino teachers”
As a practical matter, the reality is that many of today’s online English conversation services for children employ teachers based in the Philippines. The low prices are largely due to differences in labor costs. Here is a frank summary of how to think about this.
In research terms, the perspective is simple. “Whether the teacher can respond to the child’s reactions” and “whether the child can enjoy the relationship with that teacher” are far more important than whether the teacher is a native or non-native speaker. This is as summarized in the sister article How should you choose an English conversation school?, and it is the same whether in person or online.
However, there are some practical points to keep in mind.
- Filipino English may include sound qualities somewhat different from the so-called “standard pronunciation” of American or British English. However, there is no research basis that this puts young children at a decisive disadvantage in acquiring English
- Whether a teacher has plenty of experience teaching children varies greatly by operating company and individual teacher. It is realistic to choose services or teachers that clearly state they are “specialized in children” or “kids specialists”
- Whether you can keep requesting the same teacher is also important. With a different teacher every time, the relationship does not build up, and the child’s sense of security is slow to develop
“Cheap = low quality” is not true, nor is “native speaker = absolutely good”. Using “experience with children” and “whether an ongoing relationship can be built” as your yardstick makes it easier to compare regardless of price range.
Development of the first language (Japanese) and CALP
If you try to do online English conversation every day, it takes up that much of the time for Japanese at home. Here we add one more perspective that is important in research terms.
A classic framework in second language education research is the distinction between BICS (Basic Interpersonal Communicative Skills) and CALP (Cognitive Academic Language Proficiency) .
- BICS: the ability for everyday conversation. The level of greeting, saying your favorite food, answering simple questions. Acquired relatively quickly (said to take roughly 1 to 2 years)
- CALP: language ability at the level used for thinking about abstract things, explaining in writing, and learning school subjects. Said to take 5 to 7 years or more to acquire
What Cummins has repeatedly emphasized is that “if you try to develop CALP in a second language while CALP in the first language is weak, both tend to end up half-developed”. Conversely, the research consensus is that when CALP in the first language is well developed, it tends to transfer as a foundation for CALP in the second language.
What matters in early childhood is conversation at home, reading aloud, and picture book time. These build the foundation of future CALP in Japanese. Doing online English conversation every day to the point that evening picture book time and family conversation disappear is not desirable in the long run.
As a practical guide, it is safe to
- keep online English conversation to about 15 to 25 minutes a day
- not reduce time for reading aloud and conversation in Japanese at home
- note that “restricting Japanese for the sake of English” (such as speaking to the child only in English at home too) is not recommended
Please think of the first and second languages not as competing with each other but as being in the relationship of a foundation and the structure built on it.
10 to 25 minutes every day vs. 1 hour once a week: which works better?
One of the biggest advantages of online English conversation is that it is easy to set up a schedule of “short and every day”. This also makes sense in research terms.
As also mentioned in the sister article, human memory and learning are strengthened by spaced practice. If you secure the same total time, comparing
- once a week × 60 minutes (massed)
- every day × 10 to 15 minutes (spaced)
a wide range of research in educational psychology shows that the spaced approach has the advantage in terms of retention. Pinter (2017) also states that in young children’s language learning, “short, fun, and frequent” is the principle.
However, this does not mean “it is pointless unless you do it every day”. Even 3 to 4 times a week at about 15 minutes per session adds up well enough, as long as the pace is manageable for the child. “Increasing the frequency to the point that the child resists” is counterproductive, and this connects to point (3), “whether the child can enjoy it”.
The idea of a family mix
Another possibility of online English conversation is that it is easy to share as a family. Unlike going to an in-person school, because you do it at home, you can use it in ways like these.
- Siblings share the time (20 minutes for the older child, 10 minutes for the younger)
- The parent also listens from behind, and after the lesson you recall things together: “What was that again?”
- The parent takes separate lessons for themselves too, showing the child that “Mom is working hard too”
This is not just a money-saving trick; it is also reasonable in research terms, in the sense of creating an atmosphere in which “English is part of the family’s daily life”. Second language acquisition research has repeatedly pointed out that the home language environment (how the family is involved in language learning) has a large influence on children’s motivation and persistence (Pinter, 2017).
Conversely, with a style of “only the child does it” and “the parent stays out of it”, the child tends to feel alone and is less likely to keep going. It risks becoming nothing more than paying money to sit the child in front of a screen.
Relationship with public education: its place relative to foreign language in elementary school
Finally, a word on the relationship with public education.
The National Curriculum Guidelines for Elementary Schools (issued 2017): Foreign Languages set “Foreign Language Activities” (one period a week) in grades 3 and 4 of elementary school, and “Foreign Language” as a subject (two periods a week) in grades 5 and 6. In other words, under the current system, a certain amount of contact with English within public education begins from grade 3 of elementary school.
This does not at all mean that “children must do online English conversation before entering school”. Rather the opposite: the realistic position of online English conversation in the preschool years is to aim for effects such as
- creating a state in which the child has no resistance to English before it starts in public education
- having the feeling that “talking with adults from other countries isn’t scary”
If you set the goal not as “making them fluent” or “getting them to learn ahead” but as “keeping their relationship with English positive”, it is consistent with the research findings and worth the cost to the household budget.
Closing dialogue: choose by “two-way interaction, amount of conversation, and concentration”
So in the end, is it all right for my child to do online English conversation?
Rather than “it’s all right”, the answer from research is that “if three conditions are met, it’s a strong option”. The three are: (1) it is real-time and two-way, (2) there is an amount of conversation that lets the child respond, and (3) it fits within the length your child can stay focused for (about 15 to 25 minutes).
I have both the feeling that I want my child to do it every day because it’s cheap, and the worry that my child will get tired…
I understand. But making it every day “because it’s cheap” often leads to the child getting fed up, and it doesn’t last. Start with about 3 to 4 times a week, 15 to 20 minutes each, and if your child says “I want to do it tomorrow too”, increase it little by little. This pace is realistic and makes sense in research terms too.
I’ve also always felt guilty about sitting my child in front of a screen. Like, “our screen time ends up being long…”
That’s an area where research has been updated too. The American Academy of Pediatrics also states explicitly that “video chat is a separate category”. If your child is taking part in responsive exchanges, you can see it not as “showing a screen” but as “talking with a person through a screen”. There’s no need to feel guilty.
Also, I thought I had to choose either an in-person school or online, but is it fine to combine them?
Of course. In fact, what many researchers would likely support is a combination of “creating experiences of interacting directly with people at an in-person school, while adding frequency at home in a spaced way online”. If your budget and time allow, it’s one good design. If not, even one or the other has plenty of meaning as long as you keep the three conditions in mind.
Without deciding that “screens = bad”, choose by the three: two-way interaction, amount of conversation, and concentration. With that, I think I can decide while watching how my daughter is doing.
If you also read the sister articles How should you choose an English conversation school?, Does early English education really work?, and Phonics: when should you start?, I think it will be easier to set your family’s policy along three axes: in person, online, and reading and writing.
Research details
Primary sourcesStudy design: An experimental study assigning infants aged 9 to 10 months in an English-speaking environment to three groups + a control group
Participants: A total of 32 infants growing up in an English-speaking environment in the US (a group interacting with a live Chinese speaker, a group watching video of a Chinese speaker, and a group hearing only Chinese audio) + an English-only control group with no exposure to Chinese
Main results: After 4 weeks (12 sessions in total) of exposure, only the group that interacted with a live Chinese speaker became able to discriminate phonemes specific to Chinese. The video group and the audio-only group did not differ from the control group with no exposure to Chinese at all (differences between groups significant at the p<0.05 level). The authors theorized this as the “social gating hypothesis”.
Limitations: A study of infants; care is needed in generalizing directly to early childhood and beyond. The sample size is also relatively small. However, the finding that “recorded video alone has little language learning effect” has been replicated repeatedly in many later studies.
Study design: An experimental study comparing new verb learning by assigning children aged 24 to 30 months to three groups: (A) a face-to-face interaction group, (B) a video chat (real-time, two-way) group, and (C) a recorded video viewing group
Participants: Toddlers aged 24 to 30 months growing up in an English-speaking environment in the US
Main results: The face-to-face group and the video chat group were able to learn the new verbs to the same degree, while the recorded video group showed almost no learning effect (differences between groups significant at p<0.05). The authors interpreted this as showing the importance of “socially contingent interaction”. Consistent with Kuhl’s social gating hypothesis, they argue that when a child’s brain detects that “the other person is responding to me”, it opens the gate to deep processing.
Limitations: A limited indicator: short-term new verb learning in a laboratory. Effects on long-term language development need separate verification. However, the finding that “if it is real-time and two-way, learning effects close to face-to-face interaction appear even through a screen” is widely referred to in modern language education and developmental research.
Study design: A comprehensive review of research and practice in language teaching for young children and children (Teaching Young Language Learners) (2nd ed.)
Scope: Integration of existing research on second language learners from preschool to elementary school age
Main results: In young children’s second language learning, “short, fun, and frequent” is the principle. Motivation, affective factors, and the home language environment strongly influence learning outcomes. The relationship between the teacher and the child, the amount of opportunity for the child to respond, and the design of the length of each lesson are said to make the difference in long-term persistence and effect.
Limitations: A comprehensive review, not a rigorous measurement of the effects of a specific intervention program. However, it is widely adopted internationally as a standard reference in the field of young learners’ language education.
Study design: Policy Statement by the American Academy of Pediatrics
Scope: Integration of research findings and clinical guidance on media use by children aged 0 to 5
Main results: Recommends that under 18 months, use of digital media be avoided except for video chat; at 18 to 24 months, high-quality programs together with a parent; and at ages 2 to 5, no more than 1 hour a day of high-quality programs. The point that it explicitly positions “video chat as an exception” is important, and is consistent with findings such as the Roseberry study that “two-way video chat has effects close to face-to-face interaction”.
Limitations: A policy statement that does not present new data. Whether it applies directly to the Japanese context needs separate consideration, but the idea of distinguishing video chat from mere screen time is widely referred to.
Study design: A conceptual and empirical review of BICS (everyday conversational ability) and CALP (cognitive academic language proficiency)
Scope: More than 30 years of accumulated research on bilingual education and second language acquisition
Main results: BICS (everyday conversational ability) is acquired in 1 to 2 years, but CALP (the language ability that supports learning school subjects and abstract thinking) takes 5 to 7 years or more to develop to a native-like level. CALP being well developed in the first language becomes the foundation for CALP development in the second language (cross-linguistic transfer). Conversely, it is suggested that leaning toward the second language while CALP in the first language is weak tends to leave both half-developed.
Limitations: It is strongly positioned as a conceptual model, and the same number of years does not apply to every language and context. However, the practical guideline of “not squeezing first language development” is widely shared in the SLA field.
Study design: Official notice (issued 2017, fully implemented from FY2020)
Scope: The institutional design of foreign language (English) education in elementary schools nationwide
Main results: Sets “Foreign Language Activities” (one period a week) in grades 3 and 4 and “Foreign Language” as a subject (two periods a week) in grades 5 and 6 of elementary school. The goals are “the groundwork for communication” and “the basics of communication”, centered on listening and speaking. The axis is not “aiming for fluency” but getting familiar with foreign languages and developing a positive attitude toward communication.
Limitations: An institutional design, not an evidence-based intervention study. However, it makes possible a realistic positioning of online English conversation in the preschool years, based on “what starts in public education”.