Over the same period, determine whether there is any difference

Over the same period, determine whether there is any difference in the weight change trajectory for babies who meet their nutritional goals versus babies who do not meet their nutritional goals (all goals).

1.

Now that we've answered the question about whether there were weight changes for all subjects (and then changes during the two periods), we want to know whether there were any differences between the two comparison groups (all nutrition goals met versus not met) for the same periods (overall, birth to 28 days, and 28 days to discharge). 

Formulate three null hypotheses to reflect these new questions

2.

Now let's look more closely at the separate periods.

As it turns out, the weight change trajectories were not significantly different between the two groups of babies in all periods.

During which period do we see significant differences in weight change trajectories between the two nutritional groups?

What is the F statistic for the difference in weight change trajectories of the two groups during the period in which the trajectories were significantly different? Answer What is the p value? Answer

What is the change in the means of the two groups during this period:

• for the babies who met all their nutritional goals? Answer grams

• for the babies who did NOT meet all their nutritional goals? Answer grams

3.

Report and interpret the findings with respect to the difference in weight changes for the two groups for the three hypotheses above

Though we've determined that there is a significant difference between nutrition groups in terms of weight change, we notice that the two groups are different in terms of the length of stay. So, we wonder whether our previous findings might be altered if we take NICU length of stay (LOS) into account.

Now, we take the same model we built in the above questions, add length of stay as a covariate.

4. After controlling for length of stay, we see that the difference in the overall weight change trajectories between the two groups has changed.

What is the F statistic for the overall weight change trajectory? Answer

What is the p value? Answer

Which of the following is a reasonable conclusion regarding our hypothesis that there is a difference in weight change trajectories of the two nutritional groups after controlling for length of stay?

We fail to reject the null and conclude that there is no difference in weight change trajectories of babies who did and who did not meet their nutritional goals after controlling for length of stay

We reject the null and conclude that there is no difference in weight change trajectories of babies who did and who did not meet their nutritional goals after controlling for length of stay

We reject the null and conclude that even after controlling for length of stay, babies who met their nutritional goals still had significantly different weight change trajectories than babies who did not meet their nutritional goals

5. We would like some explanation for why the results changed as dramatically as they did before and after we controlled for length of stay.

First, what is the mean number of days that babies who met all their nutritional goals stayed in the NICU? Answer

What is the mean number of days that babies who did NOT meet all their nutritional goals stayed in the NICU? Answer

What proportion of the variance is explained by length of stay?

54.3%

84%

.3%

78.3%

1%

6.

Briefly, why might controlling for LOS change the findings?

7.