The sentence corpus was composed of regular sentences and proverbs (Fernández et al., 2013a [10] for a Corpus description). Sentences comprise a well-balanced number of content and function words, and had similar grammatical structure. We used the Spanish Lexical Léxesp corpus [11] for assigning a frequency to each word of the sentence corpus. Word predictability was measured in an independent experiment with 18 researchers of the Electrical Engineering and Computer Science Department of Universidad Nacional del Sur. We used an incremental cloze task procedure in which participants had to guess the next word given only the prior words of the sentence [10].

LMM are linear models in which the linear predictor is contained in addition to the usual fixed effects. The LMM makes it possible to account for the correlation within profiles and to consider the profiles as a random sample from a common population distribution, which is, generally, more realistic. We used the lmer program of the lme4 package (version 0.999999–2) [12] for estimating fixed and random coefficients. We chose log gaze duration as the dependent variable because this measure includes refixations on a word, and refixations usually reflect a lexical-processing difficulty for word N [3, 4, 13]. Fixed effects in LMM terminology correspond to regression coefficients in standard linear regression models. They can also estimate slopes or differences between conditions. A number of fixed effects were entered into the model: logit predictabilities (i.e., the average predictability measured from the cloze task transformed using a logit function [10]), log frequencies and 1/length of word N − 1, word N, and word N + 1. Using the reciprocal of word length (i.e., 1/length), renders the multiplicative interaction of frequency and length or predictability and length as a ratio or relative frequency and predictability measure (i.e., normalized on word length). Regression coefficients (bs) standard errors (SEs) and t-values (t = b/SE) are reported for the LMMs. In this work, we only present a summarized graphical view of the outcomes. The interested reader can find a complete description of the LMMs results in [14]. Since there is no clear definition of “degree of freedom” for LMMs, precise p-values cannot be reported. In general, however, given the large number of observations, subjects, and items entering our analysis and the comparatively small number of fixed and random effects estimated, the t-distribution is equivalent to the normal distribution for all practical purposes (i.e., the contribution of the degrees of freedom to the test statistics is negligible). Our criterion for referring to an effect as significant is t = b/SE > ±1.96.

Early diagnosis of AD is still difficult. People with early to moderate AD usually show impairment in learning and a deterioration of episodic memory, symptoms that are typically used for diagnosis of the pathology. However, the subtle alterations in movement coordination and planning that may also be present while performing fine motor tasks such as writing or reading at the very beginning of the disease are harder to detect and go commonly unnoticed [15, 16]. Therefore, it is difficult to get an early diagnosis of this disease. Evaluation of eye movements might provide considerable insight into the integrity of control circuits in AD.

Our hypothesis was that in AD patients, an increase in average cloze predictability of the incoming word would probably not facilitate reading for impairments in top-down processing. To test this hypothesis, we evaluated the eye movements in control and AD patients during the reading of sentences with either high or low average word predictability. We investigated whether an increase in the average predictability of the upcoming word (N + 1) affected gaze duration (i.e., the sum of consecutive forward fixations on a word) in both groups. Our results [14] showed that while high predictable sentence and word predictability exerted its influence on gaze duration in healthy subjects, such predictability did not modify word processing during reading in mild AD patients.

Little is known about the effect of schizophrenia on eye movement behavior during reading sentences with different contextual predictability (e.g., proverbs vs regular sentences). Some previous studies evidenced schizophrenia-related reading difficulties [21–24]. Abnormalities in both language and oculomotor control are well documented in individuals with schizophrenia [24–28]. However, fewer studies have examined the capacities through which linguistic material is processed when reading proverbs. The same example of the eye movements recorded during reading of two sentences, showing eye fixations of both controls and AD in Fig. 21.2 is repeated for control and SZ patients in Fig. 21.5. At first glance, it can be appreciated that the main difference between SZ and control is in regular (low predictable) sentences.