eyetracking_new.knit

List of variables

Participants and trials

ID
group: TD vs DYS
class: 1-5 + uni

Design

condition: cognate vs. nonword
item_n: item number
list: list number
trial: trial number
target: target item
sentence

Accuracy and validation measures

eye_used: eye used for recording
offset_drift: average drift accuracy by trial
avg_validation: average validation accuracy by ID
center:right_bot: average validation accuracy for each point

Eye-movements

Previous saccade

previous_sac_start_x: X coordinate of the previous saccade start point
previous_sac_end_x: X coordinate of the previous saccaede end point
previous_sac_start_y: Y coordinate of the previous saccade start point
previous_sac_end_y: Y coordinate of the previous saccade end point
previous_sac_amplitude: Amplitude of the previous saccade in degrees of visual angle
previous_sac_avg_velocity: Average velocity of the previous saccade
previous_sac_peak_velocity: Peak value of gaze velocity (in visual degrees per second) of the previous saccade
previous_sac_direction: Direction (Left, Right, Top, Bottom), relative to the current fixation, in which the previous saccade is aiming. Note a missing value will be recorded if the PREVIOUS_SAC_AMPLITUDE is less than 0.5 degree
previous_sac_duration: Duration of the previous saccade in milliseconds

Previous fixation

previous_fix_x: X coordinate of the previous fixation
previous_fix_y: Y coordinate of the previous fixation
previous_fix_distance: Distance of the previous fix relative to current fix in visual degrees
previous_fix_direction: Direction (Left, Right, Top, Bottom), relative to the current fixation, in which the previous fixation is located.

Current fixation

current_fix_x_warp: X coordinate of the current fix outputed by warp algorithm
current_fix_y_warp: Y coordinate of the current fix outputed by warp algorithm
current_fix_x: X coordinate of the current fixation
current_fix_y: Y coordinate of the current fixation
current_fix_start: Time current fix starts
current_fix_end: Time current fix ends
current_fix_pupil: Average pupil size during the current fixation.
current_fix_duration: Current fixation duration
ia_first_run_landing_position: Number of pixels from the horizontal position of the first fixation during the first run on the current interest area to the left edge of the interest area.

Next saccade

next_sac_start_x: X coordinate of the next saccade start point
next_sac_end_x: X coordinate of the next saccade end point
next_sac_start_y: Y coordinate of the next saccade start point
next_sac_end_y: Y coordinate of the next saccade end point
next_sac_amplitude: Amplitude of the next saccade in degrees of visual angle
next_sac_avg_velocity: Average velocity of the next saccade
next_sac_peak_velocity: Peak value of gaze velocity (in visual degrees per second) of the next saccade
next_sac_direction: Direction (Left, Right, Top, Bottom), relative to the current fixation, in which the next saccade is aiming. Note a missing value will be recorded if the NEXT_SAC_AMPLITUDE is less than 0.5 degree.
next_sac_duration: Duration of the next saccade in milliseconds.

Next fixation

next_fix_x: X coordinate of the next fixation
next_fix_y: Y coordinate of the next fixation
next_fix_direction: Direction (Left, Right, Top, Bottom), relative to the current fixation, in which the next fixation is located.
next_fix_distance: distance of the next fix relative to current fix in visual degrees

AOI pixel parameters:

ia_id: ID interest area
ia_bottom: Y-coordinate of the lower-right corner of a rectangular AOI
ia_top: Y-coordinate of the upper-left corner of a rectangular AOI
ia_right: X-coordinate of the lower-right corner of a rectangular AOI
ia_left: X-coordinate of the upper-left corner

Measures participants

wr_time_z:nwr_error_z: reading measures from std. tests in sd deviation
pa_rt, pa_acc, pa_bis: rt, accuracy and speed-accuracy trade-off in spoonerism task (PA)
forward and backward (memory measures)
lexita_acc, lexita_bis, lexita_rt: LexITA knowledge tot. score, speed-accuracy trade-off and RT
lextale_acc, lextale_bis, lextale_rt: LexTALE knowledge tot. score, speed-accuracy trade-off and RT
it_ok_acc: orthographic knowledge tot. score IT (rt and speed-accuracy trade-off not computed as test max time was 1 min)
en_ok_acc, en_ok_rt, en_ok_bis: orthographic knowledge tot. score, RT and speed-accuracy trade-off
va_span: visual attention span - tot. score
eng_prof: self-assessed EN proficiency
eng_read: self-assessed EN reading exposure
it_read: self-assessed IT reading exposure
eng_use: self-assessed EN use

Measures items

syl1_freq_pos1: frequency of the first syllable in first position
syl1_tot_freq: frequency of the first syllable
length: target word length in letters
frequenc_.wordfreq: frequency of target word from wordfreq.3
frequency_subtlex: frequency of target word from subtlex-IT and subtlex-US
biagram_frequency: biagram frequency of target words

NEW fixations parameters:

sum.fix: sum fixations for each target and ID
sum.fp.fix: Sum first pass fixation for each target and ID
tot.fix.duration: Tot fixiation duration for each target and ID
tot.fp.fix.duration: Tot first-pass fixation duration for each target and ID
fflp: First fixation landing position (rightward/leftward) based on mean first X coordinate
fflp.2: First fixation landing position (numeric) based on mean first X coordinate
is.target.fix: whether the fixation falls within AOI
is.fp.fix: whether the fixation is a first pass fixation.

Participants’ exclusion

### original sample 
df.short %>% 
  filter(!startsWith(ID, "VER")) %>% 
  summarize(
    n = n_distinct(ID), 
    .by = "group"
  ) %>% kable(caption = "Original sample") %>% kable_minimal()

Original sample
group	n
TD	61
DYS	29

### Exclude participants 
# df.short %>% 
#   
#   ### calculate reading score for exclusion (==VAS)
#   rowwise() %>% 
#   mutate(reading.score = mean(c_across(wr_time_z:nwr_error_z)),
#          reading.score = round(reading.score,2)) %>% 
#   ungroup() %>% 
#   
#   
#   ### exclude ADHD and TD with other diagnoses + DYS with balbuzia since it's reading aloud 
#   mutate(
#     exclude = case_when(
#       ID == "VER02" ~ 1, # ADHD - university - DYS (only excluded for VA Span task)
#       ID == "VER04" ~ 1, # ADHD - university - DYS (only excluded for VA Span task)
#       ID == "MEN12" ~ 1, # discalculia - 3rd - TD
#       ID == "MEN18" ~ 1, # disgrafia, discalculia - 3rd - TD 
#       ID == "MEN26" ~ 1, # disgrafie - 5th - TD, 
#       ID == "LAE33" ~ 1, # disgrafia, disortografia, 3rd - TD 
#       ID == "MEN09" ~ 1, # balbuzia - DYS 
#       TRUE ~ 0
#     ),
#   ) %>% 
#   filter(exclude == 0) %>% 
#   ### exclude poor readers 
#   filter(group.exclusion != "PR") %>% 
#   # filter(group == "DYS" | (group == "TD" & reading.score > -2)) %>% 
#   
#   
#   ### exclude participants from VER 
#   filter(ID != "VER01" & ID != "VER03") %>% 
#   dplyr::select(-exclude, -reading.score) -> df.short

df.short %>% 
  rowwise() %>% 
   ### calculate reading score 
  mutate(reading.score = mean(c_across(wr_time_z:nwr_error_z)),
         reading.score = round(reading.score,2)) %>%
  ungroup() %>%
  filter(!startsWith(ID, "VER")) %>% 
    mutate(
    exclude = case_when(
      ID == "MEN12" ~ 1, # discalculia - 3rd - TD
      ID == "MEN18" ~ 1, # disgrafia, discalculia - 3rd - TD
      ID == "MEN26" ~ 1, # disgrafia - 5th - TD, 
      ID == "LAE33" ~ 1, # disgrafia, disortografia, 3rd - TD, 
      
      ### poor reading performance: 
      ID == "LAE38" ~ 1,
      ID == "LC19" ~ 1, 
      ID == "MEN29" ~ 1, 
      ID == "MEN02" ~ 1, 
      
      
      ID == "MEN10" ~ 1, 
      ID == "LAE23" ~ 1, 
      # ID == "MEN35" ~ 1, 
      
      ID == "MEN09" ~ 1, # balbuzia - DYS 
      
      TRUE ~ 0
    )
  ) %>%
  filter(exclude == 0) -> df.short

df.short %>% 
  summarize(
    n = n_distinct(ID), 
    .by = "group"
  ) %>% 
  kable(caption = "Analyzed sample") %>% kable_minimal()

Analyzed sample
group	n
TD	51
DYS	28

### age 
df.short %>% 
  summarize(mean = mean(age), 
            sd = sd(age), 
            range = range(age)) %>% 
  mutate(across(mean:sd, ~round(.x,2))) %>% 
  kable(caption = "age") %>% kable_minimal()

age
mean	sd	range
17.04	1.44	14.11
17.04	1.44	20.50

### age of onset 
df.short %>% 
  summarize(mean = mean(AoO), 
            sd = sd(AoO), 
            range = range(AoO)) %>% 
  mutate(across(mean:sd, ~round(.x,2))) %>% 
  kable(caption = "AoO") %>% kable_minimal()

AoO
mean	sd	range
5.68	0.87	3
5.68	0.87	8

Reading scores calculation

### create a single score for the analysis 
df.model %>% 
  
  rowwise() %>% 
  mutate(
    reading.score = mean(c_across(wr_time_z:nwr_error_z))
  ) %>% 
  ungroup() -> df.model

Preliminary measures

Word reading time
group	mean	sd	range
TD	-0.42	0.85	-2.58-1.65
DYS	-3.80	1.62	-7.74–0.58

Word reading errors
group	mean	sd	range
TD	0.09	0.78	-2.36-1.18
DYS	-2.52	2.31	-8.79-1.04

Nonword reading time
group	mean	sd	range
TD	-0.07	0.81	-1.81-1.8
DYS	-2.84	1.67	-6.92–0.44

Nonword reading errors
group	mean	sd	range
TD	-0.14	1.21	-4.15-1.29
DYS	-2.73	2.60	-8.97-0.75

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = wr_time_z ~ group) -> m1 
tidy(m1) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Word reading time") %>% kable_minimal()

Word reading time
term	estimate	std.error	statistic	p.value
(Intercept)	-3.797	0.223	-17.032	0
groupTD	3.381	0.277	12.185	0

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = wr_error_z ~ group) -> m2
tidy(m2) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Word reading errors") %>% kable_minimal()

Word reading errors
term	estimate	std.error	statistic	p.value
(Intercept)	-2.524	0.285	-8.858	0
groupTD	2.618	0.355	7.381	0

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = nwr_time_z ~ group) -> m3
tidy(m3) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Nonword reading time") %>% kable_minimal()

Nonword reading time
term	estimate	std.error	statistic	p.value
(Intercept)	-2.841	0.224	-12.664	0
groupTD	2.771	0.279	9.926	0

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = nwr_error_z ~ group) -> m4
tidy(m4) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Nonword reading errors") %>% kable_minimal()

Nonword reading errors
term	estimate	std.error	statistic	p.value
(Intercept)	-2.727	0.344	-7.92	0
groupTD	2.588	0.428	6.04	0

PA
group	mean	sd	range
TD	0.99	0.91	-2.12-2.13
DYS	-1.30	2.01	-6.47-1.68

Italian Orthographic Knowledge
group	mean	sd	range
TD	46.35	9.36	25-59
DYS	30.18	9.78	14-50

English Orthographic Knowledge
group	mean	sd	range
TD	0.89	0.77	-0.79-2.47
DYS	-1.09	1.54	-5.62-2.3

LexITA
group	mean	sd	range
TD	57.06	2.01	48-60
DYS	52.61	8.18	17-60

LexTALE
group	mean	sd	range
TD	13.33	6.32	0-34
DYS	8.89	6.27	-2-29

Verbal STM (F)
group	mean	sd	range
TD	6.63	1.11	4-9
DYS	6.00	0.90	4-8

Verbal STM (B)
group	mean	sd	range
TD	5.39	1.40	3-8
DYS	4.29	1.92	0-8

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = pa.bis ~ age * group) -> m5
tidy(m5) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Phonological Awareness") %>% kable_minimal()

Phonological Awareness
term	estimate	std.error	statistic	p.value
(Intercept)	-14.156	3.652	-3.877	0.000
age	0.730	0.207	3.529	0.001
groupTD	15.429	4.216	3.660	0.000
age:groupTD	-0.747	0.242	-3.086	0.003

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  glm(formula = it.ok.acc ~ age + group, family = poisson(link = "log")) -> m6
tidy(m6) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Italian Orthographic Knowledge") %>% kable_minimal()

Italian Orthographic Knowledge
term	estimate	std.error	statistic
(Intercept)	2.461	0.224	11.008
age	0.054	0.012	4.293
groupTD	0.478	0.042	11.492

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  lm(formula = en.ok.bis ~ age + group) -> m7
tidy(m7) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "English Orthographic Knowledge") %>% kable_minimal()

English Orthographic Knowledge
term	estimate	std.error	statistic	p.value
(Intercept)	-4.100	1.575	-2.604	0.011
age	0.171	0.089	1.930	0.057
groupTD	2.141	0.268	7.997	0.000

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  glm(formula = lexita.acc ~ age * group, family = poisson(link = "log")) -> m8
tidy(m8) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "LexITA") %>% kable_minimal()

LexITA
term	estimate	std.error	statistic	p.value
(Intercept)	3.054	0.386	7.912	0.000
age	0.051	0.022	2.365	0.018
groupTD	0.986	0.441	2.239	0.025
age:groupTD	-0.051	0.025	-2.035	0.042

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  rowwise() %>% mutate(lextale.acc = lextale.acc + 2) %>% ungroup() %>% 
  glm(formula = lextale.acc ~ age + group, family = poisson(link = "log")) -> m9
tidy(m9) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "LexTALE") %>% kable_minimal()

LexTALE
term	estimate	std.error	statistic	p.value
(Intercept)	0.633	0.383	1.651	0.099
age	0.099	0.021	4.653	0.000
groupTD	0.431	0.070	6.161	0.000

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  glm(formula = forward ~ age + group, family = poisson(link = "log")) -> m10
tidy(m10) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Verbal STM (F)") %>% kable_minimal()

Verbal STM (F)
term	estimate	std.error	statistic	p.value
(Intercept)	1.796	0.573	3.136	0.002
age	0.000	0.032	-0.008	0.994
groupTD	0.099	0.099	1.001	0.317

df.short %>% 
  distinct(ID, .keep_all = T) %>% 
  glm(formula = backward ~ age + group, family = poisson(link = "log")) -> m11
tidy(m11) %>% mutate(across(estimate:p.value, ~round(.x,3))) %>% 
  kable(caption = "Verbal STM (B)") %>% kable_minimal()

Verbal STM (B)
term	estimate	std.error	statistic	p.value
(Intercept)	0.659	0.642	1.028	0.304
age	0.045	0.036	1.255	0.209
groupTD	0.271	0.114	2.376	0.018

Inferential analysis

First-pass fixation count

Real words

df.model %>% 
  filter(condition == "cognate") %>% 
  mutate(frequency_wordfreq = scale(frequency_wordfreq)) %>% 
  
  glmer(formula = sum.fp.fix ~ 
          
          lang * en.ok.bis * group +  

          ia_length + # (! prev.vers)
          # bigram_freq + 
          # it.ok.acc + 
          # frequency_wordfreq + # dropped for singular fit issues 
          
          (1|ID) + (1|target), ### Singular fit with slope for language 
        family = poisson(link = "log"), 
        control = glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 100000))) -> m1.real

# TEST  ************************************************************
# drop1(m1.real, test = "Chisq") 
# plot_model(m1.real, type = "diag")

# OUTPUT ************************************************************
anova_table(m1.real) -> anova.m1.real

	Estimate	SE	z.value	p.value
(Intercept)	0.867	0.053	16.405	0.000
langita	-0.290	0.074	-3.922	0.000
en.ok.bis	-0.137	0.036	-3.847	0.000
groupTD	-0.210	0.069	-3.061	0.002
ia_length	0.127	0.015	8.374	0.000
langita:en.ok.bis	0.047	0.051	0.934	0.350
langita:groupTD	0.132	0.096	1.372	0.170
en.ok.bis:groupTD	0.109	0.071	1.542	0.123
langita:en.ok.bis:groupTD	-0.039	0.099	-0.390	0.696

	Chisq	Df	p.value
(Intercept)	269.11	1	0.000
lang	15.38	1	0.000
en.ok.bis	14.80	1	0.000
group	9.37	1	0.002
ia_length	70.13	1	0.000
lang:en.ok.bis	0.87	1	0.350
lang:group	1.88	1	0.170
en.ok.bis:group	2.38	1	0.123
lang:en.ok.bis:group	0.15	1	0.696

Nonwords

df.model %>% 
  filter(condition == "nonword") %>% 
  
  glmer(formula = sum.fp.fix ~ 
          
          ### We use orthographic knowledge as a measure, and remove vocabulary knowledge in that we deal with nonwords 
          lang * en.ok.bis * group +

          ia_length +
          # bigram_freq +
          # it.ok.acc +
          
          (1 |ID) + (1|target), ### Singluar fit with slope for lang 
        family = poisson(link = "log"), 
        control = glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 100000))) -> m1.nw

# TEST  ************************************************************
# drop1(m1.nw, test = "Chisq")
# plot_model(m1.nw, type = "diag")

# OUTPUT ************************************************************
anova_table(m1.nw) -> anova.m1.nw

	Estimate	SE	z.value	p.value
(Intercept)	1.312	0.058	22.490	0.000
langita	-0.223	0.062	-3.585	0.000
en.ok.bis	-0.021	0.042	-0.493	0.622
groupTD	-0.235	0.073	-3.226	0.001
ia_length	0.127	0.022	5.739	0.000
langita:en.ok.bis	-0.016	0.043	-0.364	0.716
langita:groupTD	0.034	0.076	0.447	0.655
en.ok.bis:groupTD	-0.086	0.077	-1.119	0.263
langita:en.ok.bis:groupTD	0.180	0.080	2.245	0.025

	Chisq	Df	p.value
(Intercept)	505.81	1	0.000
lang	12.85	1	0.000
en.ok.bis	0.24	1	0.622
group	10.40	1	0.001
ia_length	32.94	1	0.000
lang:en.ok.bis	0.13	1	0.716
lang:group	0.20	1	0.655
en.ok.bis:group	1.25	1	0.263
lang:en.ok.bis:group	5.04	1	0.025

Explore whether Language differences are stronger in DYS vs. TD depending on proficiency levels (low: -1SD, high: +1SD)

emmeans(m1.nw, pairwise ~ lang | group | en.ok.bis, at = list(en.ok.bis = c(-1,1)), adjust = "bonferroni")$contrast %>% summary() %>% as.data.frame() %>% 
  mutate(across(estimate:p.value, ~round(.x,3))) -> emmeans1.m1.nw
color.me <- which(emmeans1.m1.nw$p.value <= .05)
emmeans1.m1.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

contrast	group	en.ok.bis	estimate	SE	df	z.ratio	p.value
eng - ita	DYS	-1	0.208	0.048	Inf	4.342	0.000
eng - ita	TD	-1	0.354	0.109	Inf	3.230	0.001
eng - ita	DYS	1	0.239	0.096	Inf	2.500	0.012
eng - ita	TD	1	0.025	0.055	Inf	0.463	0.644

Explore whether Group differences are stronger in IT vs. EN depending on proficiency levels (low: -1SD, high: +1SD)

emmeans(m1.nw, pairwise ~ group | lang | en.ok.bis, at = list(en.ok.bis = c(-1,1)), adjust = "bonferroni")$contrast %>% summary() %>% as.data.frame() %>% 
  mutate(across(estimate:p.value, ~round(.x,3))) -> emmeans2.m1.nw
color.me <- which(emmeans2.m1.nw$p.value <= .05)
emmeans2.m1.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

contrast	lang	en.ok.bis	estimate	SE	df	z.ratio	p.value
DYS - TD	eng	-1	0.149	0.110	Inf	1.349	0.177
DYS - TD	ita	-1	0.295	0.114	Inf	2.578	0.010
DYS - TD	eng	1	0.321	0.101	Inf	3.165	0.002
DYS - TD	ita	1	0.107	0.106	Inf	1.010	0.312

Explore whether English proficiency is stronger in IT vs. EN, depending on Group

emtrends(m1.nw, pairwise ~ lang | group, var = "en.ok.bis", adjust = "bonferroni", infer = T)$emtrends %>% as.data.frame() %>% mutate(across(en.ok.bis.trend:p.value, ~round(.x,3))) -> emtrends1.m1.nw

color.me <- which(emtrends1.m1.nw$p.value <= .05)
emtrends1.m1.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

lang	group	en.ok.bis.trend	SE	df	asymp.LCL	asymp.UCL	z.ratio	p.value
eng	DYS	-0.021	0.042	Inf	-0.102	0.061	-0.493	0.622
ita	DYS	-0.036	0.044	Inf	-0.122	0.050	-0.826	0.409
eng	TD	-0.107	0.065	Inf	-0.233	0.020	-1.650	0.099
ita	TD	0.057	0.066	Inf	-0.072	0.187	0.870	0.384

emtrends(m1.nw, pairwise ~ lang | group, var = "en.ok.bis", adjust = "bonferroni", infer = T)$contrasts %>% as.data.frame() %>% mutate(across(estimate:p.value, ~round(.x,3))) -> emtrends2.m1.nw

color.me <- which(emtrends2.m1.nw$p.value <= .05)
emtrends2.m1.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

contrast	group	estimate	SE	df	asymp.LCL	asymp.UCL	z.ratio	p.value
eng - ita	DYS	0.016	0.043	Inf	-0.068	0.099	0.364	0.716
eng - ita	TD	-0.164	0.068	Inf	-0.297	-0.031	-2.425	0.015

Gaze duration

Definition: Duration of first-pass fixations

Real words

df.model %>% 
  filter(condition == "cognate") %>% 
  mutate(tot.fp.fix.duration = log(tot.fp.fix.duration),
         frequency_wordfreq = scale(frequency_wordfreq)) %>% 
  
  lmer(formula = tot.fp.fix.duration ~ 
          
         lang * en.ok.bis * group +

          ia_length + 
          # bigram_freq + 
          # it.ok.acc + 
          frequency_wordfreq + 
          
          (1|ID) + (1|target) ### Singular fit with slope for lang 
       ) -> m2.real

# TEST  ************************************************************
# drop1(m2.real, test = "Chisq")
# plot_model(m2.real, type = "diag")


# OUTPUT ************************************************************
anova_table(m2.real) -> anova.m2.real

	Estimate	SE	df	t.value	p.value
(Intercept)	6.343	0.063	165.777	99.990	0.000
langita	-0.314	0.072	245.222	-4.332	0.000
en.ok.bis	-0.190	0.043	155.125	-4.448	0.000
groupTD	-0.122	0.073	152.260	-1.663	0.098
ia_length	0.092	0.021	43.276	4.342	0.000
frequency_wordfreq	-0.069	0.025	43.494	-2.783	0.008
langita:en.ok.bis	0.107	0.046	1636.195	2.307	0.021
langita:groupTD	0.059	0.079	1638.054	0.752	0.452
en.ok.bis:groupTD	0.055	0.075	144.363	0.736	0.463
langita:en.ok.bis:groupTD	0.024	0.080	1628.118	0.301	0.764

	Chisq	Df	p.value
(Intercept)	9997.99	1	0.000
lang	18.77	1	0.000
en.ok.bis	19.78	1	0.000
group	2.77	1	0.096
ia_length	18.85	1	0.000
frequency_wordfreq	7.74	1	0.005
lang:en.ok.bis	5.32	1	0.021
lang:group	0.57	1	0.452
en.ok.bis:group	0.54	1	0.462
lang:en.ok.bis:group	0.09	1	0.764

Explore whether English proficiency is stronger in IT vs. EN

emtrends(m2.real, pairwise ~ lang | group, var = "en.ok.bis", adjust = "bonferroni", infer = T)$emtrends %>% 
  as.data.frame() %>% mutate(across(en.ok.bis.trend:p.value, ~round(.x,3))) -> emtrends1.m2.real

color.me <- which(emtrends1.m2.real$p.value <= .05)
emtrends1.m2.real %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

lang	group	en.ok.bis.trend	SE	df	lower.CL	upper.CL	t.ratio	p.value
eng	DYS	-0.190	0.043	153.641	-0.275	-0.106	-4.447	0.000
ita	DYS	-0.084	0.042	148.988	-0.168	0.000	-1.975	0.050
eng	TD	-0.135	0.061	138.139	-0.256	-0.014	-2.213	0.029
ita	TD	-0.005	0.062	144.114	-0.127	0.118	-0.075	0.941

Explore language differences by proficiency levels

contrast	en.ok.bis	group	estimate	SE	df	t.ratio	p.value
eng - ita	-2	DYS	0.53	0.08	327.03	6.71	0.000
eng - ita	2	DYS	0.10	0.15	1350.81	0.69	0.491
eng - ita	-2	TD	0.52	0.17	1532.54	3.00	0.003
eng - ita	2	TD	-0.01	0.11	899.76	-0.06	0.951

Nonwords

df.model %>% 
  filter(condition == "nonword") %>% 
  mutate(tot.fp.fix.duration = log(tot.fp.fix.duration)) %>% 
  
  lmer(formula = tot.fp.fix.duration ~ 
          
          lang * en.ok.bis * group +

          ia_length + 
          # it.ok.acc + 
          # bigram_freq + 
          
          (1 |ID) + (1|target) # singular fit with slope for language 
        ) -> m2.nw

# TEST  ************************************************************
# drop1(m2.nw, test = "Chisq")
# plot_model(m2.nw, type = "diag")

# OUTPUT ************************************************************
anova_table(m2.nw) -> anova.m2.nw

	Estimate	SE	df	t.value	p.value
(Intercept)	6.992	0.085	156.968	82.052	0.000
langita	-0.311	0.096	258.644	-3.248	0.001
en.ok.bis	-0.009	0.058	138.228	-0.153	0.879
groupTD	-0.175	0.100	141.075	-1.757	0.081
ia_length	0.105	0.037	44.470	2.801	0.008
langita:en.ok.bis	-0.012	0.061	1666.278	-0.202	0.840
langita:groupTD	0.049	0.105	1669.411	0.465	0.642
en.ok.bis:groupTD	-0.209	0.102	136.648	-2.042	0.043
langita:en.ok.bis:groupTD	0.234	0.107	1660.252	2.188	0.029

	Chisq	Df	p.value
(Intercept)	6732.51	1	0.000
lang	10.55	1	0.001
en.ok.bis	0.02	1	0.879
group	3.09	1	0.079
ia_length	7.85	1	0.005
lang:en.ok.bis	0.04	1	0.840
lang:group	0.22	1	0.642
en.ok.bis:group	4.17	1	0.041
lang:en.ok.bis:group	4.79	1	0.029

Explore whether Language differences are stronger in DYS vs. TD depending on proficiency levels (low: -1SD, high: +1SD)

emmeans(m2.nw, pairwise ~ lang | group | en.ok.bis, at = list(en.ok.bis = c(-1,1)), adjust = "bonferroni")$contrast %>% summary() %>% as.data.frame() %>% 
  mutate(across(estimate:p.value, ~round(.x,3))) -> emmeans1.m2.nw
color.me <- which(emmeans1.m2.nw$p.value <= .05)
emmeans1.m2.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

contrast	group	en.ok.bis	estimate	SE	df	t.ratio	p.value
eng - ita	DYS	-1	0.299	0.079	125.295	3.795	0.000
eng - ita	TD	-1	0.483	0.148	942.360	3.276	0.001
eng - ita	DYS	1	0.323	0.140	822.417	2.315	0.021
eng - ita	TD	1	0.041	0.080	134.217	0.507	0.613

Explore whether Group differences are stronger in IT vs. EN depending on proficiency levels (low: -1SD, high: +1SD)

emmeans(m2.nw, pairwise ~ group | lang | en.ok.bis, at = list(en.ok.bis = c(-1,1)), adjust = "bonferroni")$contrast %>% summary() %>% as.data.frame() %>% 
  mutate(across(estimate:p.value, ~round(.x,3))) -> emmeans2.m2.nw
color.me <- which(emmeans2.m2.nw$p.value <= .05)
emmeans2.m2.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

contrast	lang	en.ok.bis	estimate	SE	df	t.ratio	p.value
DYS - TD	eng	-1	-0.034	0.147	137.551	-0.231	0.818
DYS - TD	ita	-1	0.151	0.147	136.916	1.027	0.306
DYS - TD	eng	1	0.384	0.139	141.190	2.773	0.006
DYS - TD	ita	1	0.102	0.140	146.814	0.727	0.469

Explore whether English proficiency is stronger in IT vs. EN, depending on Group

emtrends(m2.nw, pairwise ~ lang | group, var = "en.ok.bis", adjust = "bonferroni", infer = T)$emtrends %>% 
  as.data.frame() %>% mutate(across(en.ok.bis.trend:p.value, ~round(.x,3))) -> emtrends1.m2.nw

color.me <- which(emtrends1.m2.nw$p.value <= .05)
emtrends1.m2.nw %>% kable() %>% kable_minimal() %>% row_spec(color.me, bold = T)

lang	group	en.ok.bis.trend	SE	df	lower.CL	upper.CL	t.ratio	p.value
eng	DYS	-0.009	0.058	138.684	-0.123	0.105	-0.153	0.879
ita	DYS	-0.021	0.058	142.882	-0.136	0.094	-0.362	0.718
eng	TD	-0.218	0.085	136.449	-0.385	-0.051	-2.576	0.011
ita	TD	0.004	0.085	136.810	-0.164	0.171	0.042	0.967

contrast	group	estimate	SE	df	lower.CL	upper.CL	t.ratio	p.value
eng - ita	DYS	0.012	0.061	1665.261	-0.107	0.131	0.202	0.840
eng - ita	TD	-0.221	0.088	1657.845	-0.394	-0.049	-2.517	0.012

## To cite R in publications use:
## 
##   R Core Team (2023). _R: A Language and Environment for Statistical
##   Computing_. R Foundation for Statistical Computing, Vienna, Austria.
##   <https://www.R-project.org/>.
## 
## A BibTeX entry for LaTeX users is
## 
##   @Manual{,
##     title = {R: A Language and Environment for Statistical Computing},
##     author = {{R Core Team}},
##     organization = {R Foundation for Statistical Computing},
##     address = {Vienna, Austria},
##     year = {2023},
##     url = {https://www.R-project.org/},
##   }
## 
## We have invested a lot of time and effort in creating R, please cite it
## when using it for data analysis. See also 'citation("pkgname")' for
## citing R packages.

Other eye-tracking measures

First fixation duration

Definition: Duration of the first fixation on the target word

Real words

df.model %>% 
  filter(condition == "cognate") %>% 
  mutate(current_fix_duration = log(current_fix_duration), 
         frequency_wordfreq = scale(frequency_wordfreq)) %>% 
  
  lmer(formula = current_fix_duration ~ 
          
         lang * en.ok.bis * group +
         ia_length + 
         # frequency_wordfreq + 
         # it.ok.acc + 
          
          (1 |ID) + (1|target) # singular fit with slope for language 
        ) -> mod.first.fix.real

# TEST  ************************************************************
# drop1(mod.first.fix.real, test = "Chisq")
# plot_model(mod.first.fix.real, type = "diag")

	Chisq	Df	p.value
(Intercept)	14569.37	1	0.000
lang	0.12	1	0.732
en.ok.bis	2.79	1	0.095
group	3.98	1	0.046
ia_length	0.12	1	0.733
lang:en.ok.bis	1.88	1	0.171
lang:group	3.57	1	0.059
en.ok.bis:group	3.85	1	0.050
lang:en.ok.bis:group	1.10	1	0.294

	Estimate	SE	df	t.value	p.value
(Intercept)	5.615	0.047	179.075	120.704	0.000
langita	-0.019	0.056	328.979	-0.343	0.732
en.ok.bis	-0.054	0.032	179.188	-1.671	0.096
groupTD	0.109	0.055	175.024	1.996	0.047
ia_length	-0.005	0.014	42.322	-0.341	0.735
langita:en.ok.bis	0.051	0.037	1638.858	1.370	0.171
langita:groupTD	-0.121	0.064	1640.988	-1.890	0.059
en.ok.bis:groupTD	-0.109	0.056	165.189	-1.963	0.051
langita:en.ok.bis:groupTD	0.068	0.065	1628.794	1.048	0.295

Contrast	EOK	Group	Estimate	SE	t	p
eng - ita	-1	DYS	0.07	0.04	1.57	0.119
eng - ita	1	DYS	-0.03	0.08	-0.38	0.701
eng - ita	-1	TD	0.26	0.09	2.95	0.003
eng - ita	1	TD	0.02	0.05	0.45	0.657

Nonwords

df.model %>% 
  filter(condition == "nonword") %>% 
  mutate(current_fix_duration = log(current_fix_duration), 
         frequency_wordfreq = scale(frequency_wordfreq)) %>% 
  
  lmer(formula = current_fix_duration ~ 
          
         lang * en.ok.bis * group +
         ia_length + 
         # it.ok.acc + 
         
          
          (1 |ID) + (1|target) # singular fit with slope for language 
        ) -> mod.first.fix.nw

# TEST  ************************************************************
# drop1(mod.first.fix.nw, test = "Chisq")

	Chisq	Df	p.value
(Intercept)	8026.07	1	0.000
lang	1.24	1	0.265
en.ok.bis	0.06	1	0.802
group	0.16	1	0.689
ia_length	1.62	1	0.203
lang:en.ok.bis	0.08	1	0.779
lang:group	0.25	1	0.619
en.ok.bis:group	0.42	1	0.516
lang:en.ok.bis:group	0.07	1	0.798

	Estimate	SE	df	t.value	p.value
(Intercept)	5.788	0.065	134.790	89.588	0.000
langita	-0.073	0.065	512.045	-1.114	0.266
en.ok.bis	-0.012	0.047	128.368	-0.250	0.803
groupTD	0.032	0.080	130.779	0.400	0.690
ia_length	-0.027	0.021	44.808	-1.273	0.210
langita:en.ok.bis	0.013	0.046	1674.304	0.280	0.779
langita:groupTD	0.039	0.079	1677.057	0.497	0.619
en.ok.bis:groupTD	-0.054	0.083	127.270	-0.649	0.518
langita:en.ok.bis:groupTD	0.021	0.081	1662.895	0.256	0.798

Total fixation count

Definition: Total amount of fixations on the target word

Real words

df.model %>% 
  filter(condition == "cognate") %>% 
  
glmer(formula = sum.fix ~ 
          
          lang * en.ok.bis * group +

          ia_length + 
          # it.ok.acc + 
          frequency_wordfreq +
          
          (1|ID) + (1|target), ### Singular fit with slope for language 
        family = poisson(link = "log"), 
        control = glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 100000))) -> mod.tot.fix.real

# TEST  ************************************************************
# drop1(mod.tot.fix.real, test = "Chisq")
# plot_model(mod.tot.fix.real, type = "diag")

	Chisq	Df	p.value
(Intercept)	174.82	1	0.000
lang	11.15	1	0.001
en.ok.bis	13.56	1	0.000
group	14.28	1	0.000
ia_length	32.85	1	0.000
frequency_wordfreq	4.95	1	0.026
lang:en.ok.bis	0.22	1	0.639
lang:group	2.62	1	0.106
en.ok.bis:group	4.09	1	0.043
lang:en.ok.bis:group	0.74	1	0.391

	Estimate	SE	z.value	p.value
(Intercept)	1.345	0.102	13.222	0.000
langita	-0.235	0.071	-3.339	0.001
en.ok.bis	-0.149	0.040	-3.683	0.000
groupTD	-0.280	0.074	-3.779	0.000
ia_length	0.113	0.020	5.732	0.000
frequency_wordfreq	-0.198	0.089	-2.225	0.026
langita:en.ok.bis	0.021	0.044	0.469	0.639
langita:groupTD	0.135	0.083	1.618	0.106
en.ok.bis:groupTD	0.154	0.076	2.022	0.043
langita:en.ok.bis:groupTD	-0.074	0.086	-0.858	0.391

Nonwords

df.model %>% 
  filter(condition == "nonword") %>% 
  
glmer(formula = sum.fix ~ 
          
          lang * en.ok.bis * group +

          ia_length + 
          # bigram_freq + 
          # it.ok.acc +
          
          (1|ID) + (1|target), ### Singular fit with slope for language 
        family = poisson(link = "log"), 
        control = glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 100000))) -> mod.tot.fix.nw

# TEST  ************************************************************
# drop1(mod.tot.fix.nw, test = "Chisq")
# plot_model(mod.tot.fix.nw, type = "diag")

	Chisq	Df	p.value
(Intercept)	661.29	1	0.000
lang	0.76	1	0.383
en.ok.bis	2.24	1	0.134
group	16.51	1	0.000
ia_length	32.63	1	0.000
lang:en.ok.bis	3.40	1	0.065
lang:group	2.99	1	0.084
en.ok.bis:group	0.23	1	0.629
lang:en.ok.bis:group	1.06	1	0.303

	Estimate	SE	z.value	p.value
(Intercept)	1.627	0.063	25.716	0.000
langita	-0.047	0.053	-0.873	0.383
en.ok.bis	-0.067	0.045	-1.498	0.134
groupTD	-0.319	0.079	-4.063	0.000
ia_length	0.119	0.021	5.712	0.000
langita:en.ok.bis	-0.061	0.033	-1.843	0.065
langita:groupTD	0.108	0.062	1.730	0.084
en.ok.bis:groupTD	0.040	0.082	0.483	0.629
langita:en.ok.bis:groupTD	0.067	0.065	1.029	0.303

Total fixation duration

Definition: Total time spent on the target word (including gaze duration and regression on the target IA)

Real words

df.model %>% 
  filter(condition == "cognate") %>% 
  mutate(composite.en = scale(composite.en)) %>% 
  mutate(tot.fix.duration = log(tot.fix.duration),
         frequency_wordfreq = scale(frequency_wordfreq)) %>% 
  
  lmer(formula = tot.fix.duration ~ 
          
          lang * en.ok.bis * group +

          ia_length + 
          # bigram_freq + 
          # it.ok.acc + 
          frequency_wordfreq + 
          
          (1|ID) + (1|target) ### Singular fit with slope for lang 
       ) -> mod.fix.dur.real

# TEST  ************************************************************
# drop1(mod.fix.dur.real, test = "Chisq")
# plot_model(mod.fix.dur.real, type = "diag")

	Chisq	Df	p.value
(Intercept)	10704.28	1	0.000
lang	24.12	1	0.000
en.ok.bis	30.85	1	0.000
group	16.69	1	0.000
ia_length	20.37	1	0.000
frequency_wordfreq	5.63	1	0.018
lang:en.ok.bis	5.00	1	0.025
lang:group	5.00	1	0.025
en.ok.bis:group	2.98	1	0.085
lang:en.ok.bis:group	0.48	1	0.486

	Estimate	SE	df	t.value	p.value
(Intercept)	6.734	0.065	147.072	103.462	0.000
langita	-0.342	0.070	129.002	-4.911	0.000
en.ok.bis	-0.229	0.041	124.122	-5.555	0.000
groupTD	-0.288	0.071	122.665	-4.085	0.000
ia_length	0.109	0.024	43.600	4.513	0.000
frequency_wordfreq	-0.068	0.028	43.903	-2.374	0.022
langita:en.ok.bis	0.084	0.038	1629.288	2.236	0.025
langita:groupTD	0.143	0.064	1630.743	2.235	0.026
en.ok.bis:groupTD	0.125	0.072	117.582	1.725	0.087
langita:en.ok.bis:groupTD	-0.046	0.065	1624.861	-0.696	0.486

contrast	group	en.ok.bis	estimate	SE	df	t.ratio	p.value
eng - ita	DYS	-1	0.43	0.06	77.75	6.96	0.000
eng - ita	TD	-1	0.24	0.10	433.87	2.43	0.015
eng - ita	DYS	1	0.26	0.09	379.74	2.75	0.006
eng - ita	TD	1	0.16	0.06	80.50	2.59	0.011

contrast	lang	en.ok.bis	estimate	SE	df	t.ratio	p.value
DYS - TD	eng	-1	0.41	0.1	113.76	3.99	0.000
DYS - TD	ita	-1	0.22	0.1	116.11	2.15	0.033
DYS - TD	eng	1	0.16	0.1	122.19	1.66	0.100
DYS - TD	ita	1	0.07	0.1	119.08	0.67	0.504

Nonwords

	Chisq	Df	p.value
(Intercept)	11599.24	1	0.000
lang	0.99	1	0.319
en.ok.bis	11.67	1	0.001
group	15.83	1	0.000
ia_length	10.39	1	0.001
lang:en.ok.bis	0.58	1	0.445
lang:group	0.47	1	0.492
en.ok.bis:group	0.36	1	0.547
lang:en.ok.bis:group	4.79	1	0.029

	Estimate	SE	df	t.value	p.value
(Intercept)	7.453	0.069	132.145	107.700	0.000
langita	-0.067	0.067	93.467	-0.996	0.322
en.ok.bis	-0.148	0.043	96.573	-3.417	0.001
groupTD	-0.297	0.075	97.677	-3.978	0.000
ia_length	0.111	0.034	44.196	3.224	0.002
langita:en.ok.bis	-0.024	0.031	1656.139	-0.764	0.445
langita:groupTD	0.037	0.053	1657.541	0.688	0.492
en.ok.bis:groupTD	-0.046	0.077	95.856	-0.602	0.549
langita:en.ok.bis:groupTD	0.119	0.054	1654.237	2.188	0.029

Regression path

Definition: First-pass fixation duration + duration regressions on the previous word + duration returns to IA (revisits)

### Calculate regression path 

df.long %>% 
  
  
### CONDITION 1 - Set `is.regression`
  # > if the fixation is in the IA and the direction of the next one is LEFT 
  # > if the fixation is not in the IA but the previous one was, and it was also a regression, then this is a regression too 
  # > if previous fixation is a regression, and current fix is target fixation but not a first-pass fixation, then it is a regression 
  
  mutate(
    is.regression = if_else(
  ### Condition (a)
  is.fp.fix == 1 & next_fix_direction == "LEFT", 1, 0
  ), 
  
  is.regression = if_else(
    
  ### Condition (b)
  is.fp.fix == 0 & (lag(is.fp.fix) == 1 & lag(is.regression) == 1), 1, is.regression
  ), 
  
  is.regression = if_else(
    is.target.fix == 1 & is.fp.fix == 0 & lag(is.regression) == 1, 1, is.regression
  ), 
  
is.regression = if_else(
    (is.target.fix == 1 & is.fp.fix == 0) | (lag(is.regression, default = 0) == 1 & is.target.fix == 1), 1, is.regression
)
) %>% 
  
group_by(ID, ia_label) %>% 
  mutate(
    regression_path_count = sum(is.regression, na.rm = T),
    regression_path_duration = sum(current_fix_duration[is.regression == 1], na.rm = T)
  ) %>% 
  distinct(ia_label, .keep_all = T) %>% 
ungroup() -> df.reg

# summary(df.reg$regression_path_duration)

Real words

df.reg %>% 
  filter(condition == "cognate") %>% 
  ### composite en
  relocate(lextale.acc, .after = en.ok.bis) %>% 
  mutate(
    lextale.acc = scale(lextale.acc)) %>% 
  rowwise() %>% 
  mutate(
    composite.en = mean(c_across(lextale.acc:en.ok.bis)),
    ia_length = (ia_right - ia_left)) %>% 
  ungroup() %>% 

  mutate(composite.en = scale(composite.en),
         frequency_wordfreq = scale(frequency_wordfreq),
         ia_length = scale(ia_length),
         bigram_freq = scale(bigram_freq),
         it.ok.acc = scale(it.ok.acc),
         frequency_wordfreq = scale(frequency_wordfreq),
         # regression_path_duration = log(regression_path_duration)
         ) %>% 
  
  lmer(formula = regression_path_duration ~ 
          
          lang * en.ok.bis * group + 

          ia_length + 
          # bigram_freq + 
          # it.ok.acc + 
          # frequency_wordfreq + 
          
          (1|ID) + (1|target)) -> mod.reg.real

	Chisq	Df	p.value
(Intercept)	93.76	1	0.000
lang	10.62	1	0.001
en.ok.bis	11.91	1	0.001
group	15.37	1	0.000
ia_length	6.14	1	0.013
lang:en.ok.bis	0.45	1	0.504
lang:group	5.96	1	0.015
en.ok.bis:group	0.54	1	0.464
lang:en.ok.bis:group	0.86	1	0.353

	Estimate	SE	df	t.value	p.value
(Intercept)	573.036	59.180	172.281	9.683	0.000
langita	-200.744	61.592	189.644	-3.259	0.001
en.ok.bis	-78.168	22.650	150.552	-3.451	0.001
groupTD	-259.676	66.242	152.899	-3.920	0.000
ia_length	52.641	21.251	44.395	2.477	0.017
langita:en.ok.bis	-14.120	21.117	2017.473	-0.669	0.504
langita:groupTD	151.028	61.859	2013.833	2.441	0.015
en.ok.bis:groupTD	29.514	40.338	147.111	0.732	0.466
langita:en.ok.bis:groupTD	34.863	37.517	2021.186	0.929	0.353

Nonwords

df.reg %>% 
  filter(condition == "nonword") %>% 
  ### composite en
  relocate(lextale.acc, .after = en.ok.bis) %>% 
  mutate(
    lextale.acc = scale(lextale.acc)) %>% 
  rowwise() %>% 
  mutate(
    composite.en = mean(c_across(lextale.acc:en.ok.bis)),
    ia_length = (ia_right - ia_left)) %>% 
  ungroup() %>% 

  mutate(composite.en = scale(composite.en),
         frequency_wordfreq = scale(frequency_wordfreq),
         ia_length = scale(ia_length),
         bigram_freq = scale(bigram_freq),
         it.ok.acc = scale(it.ok.acc),
         frequency_wordfreq = scale(frequency_wordfreq),
         # regression_path_duration = log(regression_path_duration)
         ) %>% 
  
  lmer(formula = regression_path_duration ~ 
          
          lang * en.ok.bis * group + 

          ia_length + 
          # bigram_freq + 
          # it.ok.acc + 
          # frequency_wordfreq + 
          
          (1|ID) + (1|target)) -> mod.reg.nw

# drop1(mod.reg.nw, test = "Chisq")

	Chisq	Df	p.value
(Intercept)	79.42	1	0.000
lang	3.55	1	0.060
en.ok.bis	15.91	1	0.000
group	10.28	1	0.001
ia_length	4.13	1	0.042
lang:en.ok.bis	0.18	1	0.669
lang:group	0.00	1	0.994
en.ok.bis:group	3.37	1	0.067
lang:en.ok.bis:group	0.86	1	0.352

	Estimate	SE	df	t.value	p.value
(Intercept)	1052.389	118.089	143.401	8.912	0.000
langita	192.910	102.373	201.393	1.884	0.061
en.ok.bis	-188.310	47.208	122.909	-3.989	0.000
groupTD	-441.439	137.657	123.377	-3.207	0.002
ia_length	71.127	35.009	45.276	2.032	0.048
langita:en.ok.bis	15.239	35.656	2008.992	0.427	0.669
langita:groupTD	0.790	104.191	2001.848	0.008	0.994
en.ok.bis:groupTD	154.861	84.420	121.907	1.834	0.069
langita:en.ok.bis:groupTD	-58.889	63.333	2012.383	-0.930	0.353

Summary

Real words

rowname	Chisq	Df	p.value
First-pass fixation count
(Intercept)	269.11	1	< .001
Language	15.38	1	< .001
EOK	14.80	1	< .001
Group	9.37	1	0.002
IA Length	70.13	1	< .001
Language * EOK	0.87	1	0.35
Language * Group	1.88	1	0.17
EOK * Group	2.38	1	0.123
Language * EOK * Group	0.15	1	0.696
Gaze duration
(Intercept)	9997.99	1	< .001
Language	18.77	1	< .001
EOK	19.78	1	< .001
Group	2.77	1	0.096
IA Length	18.85	1	< .001
Frequency	7.74	1	0.005
Language * EOK	5.32	1	0.021
Language * Group	0.57	1	0.452
EOK * Group	0.54	1	0.462
Language * EOK * Group	0.09	1	0.764
First fixation duration
(Intercept)	14569.37	1	< .001
Language	0.12	1	0.732
EOK	2.79	1	0.095
Group	3.98	1	0.046
IA Length	0.12	1	0.733
Language * EOK	1.88	1	0.171
Language * Group	3.57	1	0.059
EOK * Group	3.85	1	0.05
Language * EOK * Group	1.10	1	0.294
Total fixation count
(Intercept)	174.82	1	< .001
Language	11.15	1	0.001
EOK	13.56	1	< .001
Group	14.28	1	< .001
IA Length	32.85	1	< .001
Frequency	4.95	1	0.026
Language * EOK	0.22	1	0.639
Language * Group	2.62	1	0.106
EOK * Group	4.09	1	0.043
Language * EOK * Group	0.74	1	0.391
Total dwell time
(Intercept)	10704.28	1	< .001
Language	24.12	1	< .001
EOK	30.85	1	< .001
Group	16.69	1	< .001
IA Length	20.37	1	< .001
Frequency	5.63	1	0.018
Language * EOK	5.00	1	0.025
Language * Group	5.00	1	0.025
EOK * Group	2.98	1	0.085
Language * EOK * Group	0.48	1	0.486
Regression path duration
(Intercept)	93.76	1	< .001
Language	10.62	1	0.001
EOK	11.91	1	0.001
Group	15.37	1	< .001
IA Length	6.14	1	0.013
Language * EOK	0.45	1	0.504
Language * Group	5.96	1	0.015
EOK * Group	0.54	1	0.464
Language * EOK * Group	0.86	1	0.353

Pseudowords

rowname	Chisq	Df	p.value
First-pass fixation count
(Intercept)	505.81	1	< .001
Language	12.85	1	< .001
EOK	0.24	1	0.622
Group	10.40	1	0.001
IA Length	32.94	1	< .001
Language * EOK	0.13	1	0.716
Language * Group	0.20	1	0.655
EOK * Group	1.25	1	0.263
Language * EOK * Group	5.04	1	0.025
Gaze duration
(Intercept)	6732.51	1	< .001
Language	10.55	1	0.001
EOK	0.02	1	0.879
Group	3.09	1	0.079
IA Length	7.85	1	0.005
Language * EOK	0.04	1	0.84
Language * Group	0.22	1	0.642
EOK * Group	4.17	1	0.041
Language * EOK * Group	4.79	1	0.029
First fixation duration
(Intercept)	8026.07	1	< .001
Language	1.24	1	0.265
EOK	0.06	1	0.802
Group	0.16	1	0.689
IA Length	1.62	1	0.203
Language * EOK	0.08	1	0.779
Language * Group	0.25	1	0.619
EOK * Group	0.42	1	0.516
Language * EOK * Group	0.07	1	0.798
Total fixation count
(Intercept)	661.29	1	< .001
Language	0.76	1	0.383
EOK	2.24	1	0.134
Group	16.51	1	< .001
IA Length	32.63	1	< .001
Language * EOK	3.40	1	0.065
Language * Group	2.99	1	0.084
EOK * Group	0.23	1	0.629
Language * EOK * Group	1.06	1	0.303
Total dwell time
(Intercept)	11599.24	1	< .001
Language	0.99	1	0.319
EOK	11.67	1	0.001
Group	15.83	1	< .001
IA Length	10.39	1	0.001
Language * EOK	0.58	1	0.445
Language * Group	0.47	1	0.492
EOK * Group	0.36	1	0.547
Language * EOK * Group	4.79	1	0.029
Regression path duration
(Intercept)	79.42	1	< .001
Language	3.55	1	0.06
EOK	15.91	1	< .001
Group	10.28	1	0.001
IA Length	4.13	1	0.042
Language * EOK	0.18	1	0.669
Language * Group	0.00	1	0.994
EOK * Group	3.37	1	0.067
Language * EOK * Group	0.86	1	0.352

Eyetracking analysis

List of variables

Participants’ exclusion

Reading scores calculation

Preliminary measures

Inferential analysis

First-pass fixation count

Real words

Nonwords

Gaze duration

Real words

Nonwords

Other eye-tracking measures

First fixation duration

Real words

Nonwords

Total fixation count

Real words

Nonwords

Total fixation duration

Real words

Nonwords

Regression path

Real words

Nonwords

Summary