search

mspeekenbrink / sdam-book

Statistics: Data analysis and modelling
Other
9 stars 3 forks source link

Minor typos in Chap 12 #8

Closed dilayercelik closed 1 year ago

dilayercelik commented 1 year ago

"Fro For a given link function, [...]"

"[...] analytical solutions for maximum likelihood parameters of generalized linear models of are usually not available."

"For large-enough n, the difference between the t-distribution and the Normal distribution is so small that [...]"

"As A straightforward way to compute (approximate) confidence intervals [...]"

"For non-linear models such as generalized linear models, confidence intervals will often no not be symmetrical [...]"

"The same can be said for assuming the model model errors (residuals) follow a Normal distribution."

A typo in the following sentence? "On the log-odds log-odds or logits, every one-unit increase in X adds β1 to the log-odds."

"Figure 12.4: The Probit link function equals the standard Normal cumultive cumulative distribution function. THe The solid dark line shows the standard Normal cumulative distribution function, with arrows depicting the .05, .5, .7, and .975 quantiles."

"The main purpose of this additional example is to illustrate that other link functions which can be used."

"Poisson regression is useful for count data where the total number of possible occurrences of an event in is not given."

"For example, suppose you give participants one hour to solve as many algebra problems as they can."

"If the rate is unknown and can, in principle, take any possible value, then the there is no theoretical maximum to the number of algebra problems that can be solved in an hour."

"As we will see later in this Chapter, however, is that the overdispersion is likely due to not properly accounting for the dependencies in the data: [...]"

Incorrect rendering: "Dealing with this properly requires the use of generalized linear mixed-effects models, which we will introduce in Section @ref({sec-glmer})."

Could there be a typo in the following sentence? "When zero-count cells are present, Agresti (2018) recommends to add a value of 1/2 to all cells of saturated models (including cells with on-zero non-zero counts)."

Missing parenthesis: "Table 12.11 shows the results for a logistic mixed-effects regression model, using data from all participants of the metacognition experiment (see Section [12.5.2])(https://mspeekenbrink.github.io/sdam-book/ch-generalized-linear-models.html#sec-logistic-regression-metacognition).

mspeekenbrink commented 1 year ago

Thanks yet again, so helpful!