posted
Let me clarify, kmb: I am dubious about the particular results in this particular study and the accuracy in this particular case.
Diet and exercise in children as it relates to obesity is a lot different from what a kid is doing during the day and how it relates to learning; there are so many different factors I have doubts that they were able to accurately control for all of them and assert that one factor is causative in this particular situation.
Posts: 21182 | Registered: Sep 2004
| IP: Logged |
posted
Right. I'm just not sure why. They seem to be from a reputable university. I don't know anything other than that about this particular study, but assessing many different factors and variables is an absolutely typical part of doing this kind of research.
Posts: 11187 | Registered: Sep 2005
| IP: Logged |
posted
And let's not forget the fact that newspaper articles about scientific studies often botch a lot of facts and make unfounded claims.
Posts: 9945 | Registered: Sep 2002
| IP: Logged |
I don't really have a pony in this particulare race, I just thought I could clear up some confusion about research practices using human subjects.
Posts: 11187 | Registered: Sep 2005
| IP: Logged |
1) how much face-to-face time with your child or baby 2) how much screen time the child or baby has 3) what language developmental milestones have been reached
Actually, these were all questions on the standard First Year Well-Child checkup forms in the clinic I worked at.
It would be a simple matter to sort the data so you looked at #3 in terms of different groups based on #2, in cases matched across groups to hold #1 constant.
That is to say, #1 could easily be controlled for, and this would be a very basic thing to do for this sort of study. If it would make a differnce to someone, I would go read the cited studies and verify that this simple, reliable control was performed (or one like it). However, I am extremely busy, so unless it is of real import, I'll leave it at the possibility existing for this to have been studied well.
---
Edited to add: in the case I laid out, a claim could be made that regardless of amount of face-to-face time, the amount of screen time an infant or child has is/isn't correlated with a decrease in language skills (or whatever outcomes measure was being used). Correlation does not prove causation, but it is another link in the chain, and it does place some onus on those (in the reseaerch, treatment, and/or lay community) who disagree with causation to address what other reasons for the correlation there might be, other than causation.
----
Edited again to add: As kmboots noted, this is common social sciences research.
Posts: 14017 | Registered: May 2000
| IP: Logged |
posted
Ah, well, thank you. Mind you, even if the studies were done well and established that increased screen time was correlated with a lower rate of development, that doesn't mean that screen time isn't sometimes the best option for a given family in a given circumstance. It may well be the best thing for a particular child in a particular circumstance.
Posts: 14017 | Registered: May 2000
| IP: Logged |
posted
I am not sure that I can say my daughter doesn't get a great deal from the TV we let her watch. I do not know many 2.5 year olds that can use stethoscope correctly in a sentence. She can. And every time she comes out with something like that I ask her where did she learn that and invariably she replies "Curious George". She's learned a great deal from Brainy Baby as well. She can count to 25 and knows her ABC's. We also showed her a potty training video and it brought the concept home to her in a way we couldn't. I think TV/DVDs are fine as long as you also interact with your child and talk to them about what they see. Nothing can replace face to face time but I see nothing wrong augmenting that with good stuff like Sesame Street and other educational shows.
Posts: 601 | Registered: Sep 2002
| IP: Logged |
posted
Yes, but my daughter had been watching during that time frame and I do not think it effected her vocabulary development. If anything I think it helped it.
Posts: 601 | Registered: Sep 2002
| IP: Logged |
posted
What they didn't mention is that watching TV makes kids better at watching TV...
I know, it sounds stupid, but what are we all doing right now, if not staring into a screen? I say kids today will need the ability to watch monitors and decipher what's going on.
Who talks nowadays anyway? Text is king!
Posts: 3950 | Registered: Mar 2006
| IP: Logged |
quote:Originally posted by dawnmaria: Yes, but my daughter had been watching during that time frame and I do not think it effected her vocabulary development. If anything I think it helped it.
Ditto.
Bridget learns words watching tv that we've tried and tried to teach her for MONTHS and she doesn't learn.
Emma learned some vocab from tv but more from starfall.com when she was 14-16 months old.
Posts: 21182 | Registered: Sep 2004
| IP: Logged |
posted
There's always the possibility that babies who prefer watching TV had bad language genes to begin with. That's the way I think it worked with my anecdotal sample. Except for my one nephew who was very talkative and would get grumpy if the show were off. He was talkative, but I can't vouch for his vocabulary size. My daughter had a large vocabulary, by mysteriously disordered syntax as a baby. I think I studied language too much and screwed up my capacity to produce baby talk. Like when the baby would talk, instead of repeating back to her, I would yell "Honey, the baby produced a voiced velar trill!" Or, you know maybe her genes were messed up. She often demonstrated a reverse epenthetic insertion in her submoraic parsing.
posted
I don't hold much store in anecdotal evidence, but if we're going to try it that way it's probably best to be fair and show it the other way. My 20-month-old has been watching some TV almost every day through the entire period in question and even now, hos vocabulary is small and his usage is smaller. We've got a list of some 30 words he has said before but when I go over the list, he hasn't used 20 of them in the last 2 months. Mostly, he says the same words he's been saying for a year...hi, mom, and dad. Oh, and hot. He really likes to tell me things are hot. Unfortunately, I must have given him too good a reaction on this word because now even his cold cereal is hot. :=)
He still does a lot of communication through baby signs, which I love because otherwise I wouldn't have any idea what he wanted.
Was it TV or something else? I don't know. He's always been hugely advanced in gross and fine motor skills. he was running well before 1 year old and using a spoon without help well before 1 year, too. Lately, he's busy learning to jump and balance and put things in very small holes. (Like the Mr. Potato Head) I like to think, then, that his uninspiring verbal skills are due to the fact that he's busy doing other things but for all I know, it's the TV.
I still don't think anecdotal evidence is good, though. There's just too much that goes into a baby's development. It is known that babies will often concentrate on one skill at a time as they develop and that they set their own priorities. They set their own schedule in general.
The truth is, TV watching or no TV watching, school aged kids can talk. I hear them all the time and I couldn't tell you who watched too much TV and who didn't. I would be far, far more interested in long-range effects of TV watching than in how many words an 8-16-month-old says. In my experience, they don't say much anyway. :=)
Posts: 2392 | Registered: Sep 2005
| IP: Logged |
posted
Christine, my daughter was talking clearly at two. She sang songs with dead-on pitch and tone.
Our son barely talked. He even went to speech in kindergarten, which was sort of silly.
When he was six, he started talking, but he is still a very quiet person, now almost eleven. He was a classic "late bloomer." And he actually ponders before he answers a question(which drives me nuts. I want an answer and I want it now!)
So, studies aside, there are people who are quiet, and people who are chatterboxes, and I am not sure how any study would be able to factor in personality.
Posts: 10890 | Registered: May 2003
| IP: Logged |
posted
If they're good, and they look at a large enough group, studies can account for a lot more than anecdotal evidence can. It doesn't make for as interesting a conversation, though.
But after giving it some thought, I still insist that the vocabulary of an 8-16-month-old isn't a very useful thing to look at. Everyone assures me that receptive language is much more important at this age anyway and in due time, they'll all talk. Even the stupidest adults can talk, after all. Posts: 2392 | Registered: Sep 2005
| IP: Logged |
posted
I don't put a lot of stock into anecdotes in this case, especially with people's own children. Obvious biases aside, I sure hope there isn't any control studies being done.
Posts: 1753 | Registered: Aug 2002
| IP: Logged |
quote:Originally posted by Elizabeth: So, studies aside, there are people who are quiet, and people who are chatterboxes, and I am not sure how any study would be able to factor in personality.
This is done through randomization. Any identifiable variables are accounted for and checked to make sure the randomization worked (e.g., that approximately the same percentages of girls are in both groups, that the socioeconomic distribution is approximately the same, etc.). Sometimes there is a more specific process of stratified randomization that can be used to ensure a more equal distribution between the groups. This is a more advanced technical matter.
Unless the other sorts of characteristics that have been brought up (such as quietness or shyness) are in some way related to the way the randomization is done -- and given that this is generally via a random number generator or the like, it is highly highly unlikely -- then they should be randomly distributed across the groups, too. You are just as likely to have some shy people in Group A as Group B, especially if we know that other identifiable variables were checked to have come out randomly distributed. (That is a double-check on the randomization process itself, and so it is not necessary to ensure every variable is randomized, merely enough to show that the process was indeed random.)
Mind you, with small sample sizes, there is more likelihood that an unusual skew will come up, even if we don't see it. (If you take 6 socks out of a bag with 10 red and 10 blue socks, it might well not be an even mix of blues and reds.) However, with large sample sizes, the likelihood of this happening approaches nil. (If you take 600 socks out of a bag with 1000 red and 1000 blue socks, it may not be an exact 50%:50% mix, but it will be very very very likely to be in the ballpark of an even distribution -- say, 48-52%:48-52%.)
This is why statistics are discussed or related in terms of confidence intervals (or, another way of looking at it, the p-value). With a large enough sample size, the likelihood that characteristics unaccounted for are not randomly distributed approaces nil, as well.
And the small bit of chance left is indeed accounted for in the assessment of the numbers.
This is not intuitively obvious, but it has been studied, double-checked, and replicated over and over for more than a 100 years. It is how statistics does work int he real world, and it has been verified over and over again. Why it works is a matter that is covered in getting an advanced degree, and so it's beyond the scope of a simple thread here. Nonetheless, it does, even if that is not intuitively obvious.
Posts: 14017 | Registered: May 2000
| IP: Logged |
quote:Originally posted by Elizabeth: So, studies aside, there are people who are quiet, and people who are chatterboxes, and I am not sure how any study would be able to factor in personality.
This is done through randomization. Any identifiable variables are accounted for and checked to make sure the randomization worked (e.g., that approximately the same percentages of girls are in both groups, that the socioeconomic distribution is approximately the same, etc.). Sometimes there is a more specific process of stratified randomization that can be used to ensure a more equal distribution between the groups. This is a more advanced technical matter.
Unless the other sorts of characteristics that have been brought up (such as quietness or shyness) are in some way related to the way the randomization is done -- and given that this is generally via a random number generator or the like, it is highly highly unlikely anyway -- then they should be randomly distributed across the groups, too. You are just as likely to have some shy people in Group A as Group B, especially if we know that other identifiable variables were checked to have come out randomly distributed. (That is a double-check on the randomization process itself, and so it is not necessary to ensure every variable is randomized, merely enough to show that the process was indeed random.)
Mind you, with small sample sizes, there is more likelihood that an unusual skew will come up, even if we don't see it. (If you take 6 socks out of a bag with 10 red and 10 blue socks, it might well not be an even mix of blues and reds.) However, with large sample sizes, the likelihood of this happening approaches nil. (If you take 600 socks out of a bag with 1000 red and 1000 blue socks, it may not be an exact 50%:50% mix, but it will be very very very likely to be in the ballpark of an even distribution -- say, 48-52%:48-52%.)
This is why statistics are discussed or related in terms of confidence intervals (or, another way of looking at it, the p-value). With a large enough sample size, the likelihood that characteristics unaccounted for are not randomly distributed approaces nil, as well.
And the small bit of chance left is indeed accounted for in the assessment of the numbers.
This is not intuitively obvious, but it has been studied, double-checked, and replicated over and over for more than a hundred years. It is how statistics does work in the real world, and it has been verified over and over again. Why it works is a matter that is covered in getting an advanced degree, and so it's beyond the scope of a simple thread here. Nonetheless, it does, even if that is not intuitively obvious.
---
Edited to add: It was only during my third statistics course that I started to understand some of this. Really, it was only by working through real-life situations myself and seeing how sample size affects the analysis that I became comfortable with relying on these techniques.
posted
Probability and Statistics was one of the most useful classes I took in law school.
One of the conceptual things that helped me understand how the randomization process can be measured was the idea that we are attempting to find the probability that a result X would be obtained if the population were very different from the sample.
For example, if we have 1000 marbles, all either blue, red, or green, and pick 1 blue marble at random, we know very little about the population as a whole. We know that the probability of picking the marble is 0 if there are no blue marbles, .001 if there is 1 blue marble, and 1 if there are 1000 blue marbles.
If, however, we pick 100 blue marbles at random, we have a much better idea of the population's color characteristics as a whole. It's possible to calculate the odds that population is 10% blue, 20% blue, 100% blue, etc.
I don't know if this is useful or not, but thinking of these as probability calculations (which I have a better intuitive grasp of than I do of statistics) really helped me in that class.
Posts: 26071 | Registered: Oct 2003
| IP: Logged |
posted
Wow. My mom let us watch two hours of television when we were young (Sesame street), and we turned out okay. Me and another one of my sisters both were speaking at the age of 2... Forming sentences, not just saying 'mommy' or 'daddy'.
Of course... My brother *did* have a stutter until he was about 8 years old, and my other sister made up her own words for everything (I was 'caw-caw')
Hmmmm. Something to ponder.
Posts: 438 | Registered: May 2006
| IP: Logged |
quote:Originally posted by Javert Hugo: I don't put a lot of stock into anecdotes in this case, especially with people's own children. Obvious biases aside, I sure hope there isn't any control studies being done.
And I hope that when I am insulted in a condescending way, the insulter has the ability to form a grammatically correct insult.
Posts: 10890 | Registered: May 2003
| IP: Logged |
Printer-friendly view of this topic
![[Wink]](wink.gif)
) meter. Fair enough?
![[Smile]](smile.gif)
