It is hard for parents who have more than one child not to compare them.
Their minds like to sort and categorise - the result is often labelling (athletic or academic) or ranking (better than, stronger than, or more able in one area or another).
New research suggests that these thoughts, even if they stay unexpressed, might affect the children's academic success.
The same research suggests that these parental expectations form early, and are not affected by changes in a child's performance.
Dr Alexander Jensen, an assistant professor of human development at Brigham Young University, and Dr Susan McHale, a professor of human development at Pennsylvania State University, used data collected on pairs of siblings of either gender from 388 mostly European-American families. In families with more than two children, they focused on the eldest and second-born.
In interviews, parents were asked: "To what extent are your children different when it comes to school and the academic arena, such as getting good grades? Would you say that the younger child is a lot better at schoolwork, that the older child is a lot better at schoolwork, or are they somewhere in between?"
Some parents said there was no difference, but in cases in which parents rated the academic abilities of siblings differently, those beliefs had an effect.
If parents thought son No. 1 was better at schoolwork than son No. 2, then, as measured by grade point average (GPA), over time, that assessment tended to become more and more accurate.
On average, the one expected to do better did, increasing the spread between the two children - even after researchers controlled for things such as relative performance in the prior year and average GPAs.
As the researchers put it: "Offspring whom parents believed to be relatively more academically competent outperformed their siblings in school, and offspring whom parents believed to be relatively less competent were outperformed by their siblings."
That relative performance had another impact: "Youth who had higher GPAs relative to their siblings became comparatively more interested in academics the following year."
Dr Jensen said: "What we found was that parental beliefs play some role in relative academic performance - not the only role, and probably not even the largest role, but it's significant.
"It's not even about whether the kids are aware of it. Just the presence of distinctions was enough to make a difference."
In families where children defied expectations, those expectations did not change.
Even if the child viewed as less able outperformed a sibling with higher expectations, the parents' views (as measured in return interviews) remained "quite stable over time", the researchers wrote.
"By the time siblings are adolescents, parents might have developed firm ideas about what their children are like and how they compare - beliefs that are less malleable than youth's GPAs."
This study is far from perfect - the sample is small and geographically limited, and even the researchers would like to know more.
Dr Jensen said: "What I wish we could answer is how this is carried out. Are parents treating children differently? Do the kids pick up on the distinctions and then act on them?"
It is also unclear how these parental beliefs develop.
Gender stereotyping might have an impact, as could ordinary childhood development. Parents of both boys and girls who said there was a difference between children were more likely to identify their daughter as the more academic of the pair. Parents were also more likely to say that the older child was the stronger student.
Dr Jensen said: "An older kid at any given moment is more competent than a younger child. That doesn't mean they have greater capacity, but I think that's what kids tend to interpret it as.
"You can help each kid feel like they've got a lot of good going for them without making them feel like they're better than their siblings."
NEW YORK TIMES