Methods
Participants
A sample of convenience was recruited via social media (Facebook, Instagram) in Polish language. Overall, there were 759 complete responses; out of them, 640 met the following inclusion criteria: child's age between 2.0 and 24.9 months, delivered at term (gestational age >= 37 weeks), no older sibling with the diagnosis of autism spectrum disorder, birthweight higher than 2500g and lower than 4500g. Younger siblings of children with autism spectrum disorder are considered at elevated likelihood of developmental disorders; thus, we decided to exclude them from the analysis. Similarly, as both low (< 2500g) and high (> 4500g) birthweight may contribute to later adverse outcomes (Wang et al., 2018), we included only infants with birthweight within the normal range (see Table 1). Approximately 30% of caregivers declared that their infants were referred for pediatric physiotherapy. The most common reasons were hypertonia, hypotonia and postural asymmetry. The online questionnaire was filled in mostly by mothers (637 mothers, 3 fathers). The majority of the caregivers had completed higher education: 17 had a PhD (2.65%), 558 held a university degree (87.19%), 64 completed high school (10%) and 1 completed vocational school (0.16%). This distribution reflects national levels of education in Poland as according to Eurostat (2021) over 40% of Poles aged 25-34 years have tertiary educational attainment, whereas around 6% have less than primary, primary or lower secondary level of education. Caregivers’ subjective assessment of their family's socio-economic situation indicated that the majority of our sample belonged to the middle class: 144 (22.5%) described their economic situation as "very good," 344 (53.8%) as "good," 137 (21.4%) as "ok," 11 (1.7%) as "not good" and 2 (0.3%) as "very bad” (2 respondents, 0.3%, declined providing answer for this question).
Table 1
Sample Characteristics in Study 1.
|
Age in months
|
N (% group)
|
Mean age in months (SD)
|
Min age
|
Max age
|
Females (% group)
|
Participation in physiotherapy (% group)
|
Entire sample
|
640 (100%)
|
12.73 (5.83)
|
1.92
|
24.90
|
298 (46.6%)
|
202 (31.6%)
|
1.9-6.0
|
92 (14.38%)
|
4.40 (1.14)
|
1.92
|
5.97
|
36 (39.1%)
|
30 (32.6%)
|
6.01-12.0
|
217 (33.91%)
|
8.99 (1.76)
|
6.03
|
11.98
|
105 (48.4%)
|
69 (31.8%)
|
12.01-18.0
|
195 (30.47%)
|
15.03 (1.75)
|
12.01
|
18.00
|
93 (47.7%)
|
62 (31.8%)
|
18.01-25.0
|
136 (21.25%)
|
21.05 (1.94)
|
18.07
|
24.90
|
64 (47.1%)
|
41 (30.1%)
|
Questionnaire translation and adaptation
Permission to translate and validate the EMQ was obtained from the author of the original English version. The EMQ was translated into Polish independently by 5 authors of this paper (ZL, MS, AR, AMK, PT). All discrepancies were discussed and resolved through group consensus. Untranslatable expressions were replaced by Polish equivalents, which sometimes required providing illustrative examples. The consensus translation was reviewed by a specialist in Polish linguistics for clarity and jargon-free wording. Several mothers of young children were interviewed to assess the understanding of the instructions and all statements. The feedback was used to construct the final version of the EMQ. A professional translator conducted a back-translation (Polish to English). A comparison of the back-translation and original versions of the questionnaire was performed, and the results were used to develop the final consensus version. The EMQ-PL and back-translation were approved by the first author of the original version (Libertus & Landa, 2013).
The EMQ-PL follows the English original in the order of items. It is a list of questions that asks the parent about the presence or absence of a given infant behavior at the time of assessment. It uses a 5-point scale ranging from −2 to +2 to quantify parents’ certainty. A behavior is rated -2 if the parent is sure the child does not show the behavior yet, -1 if the child probably does not show the behavior yet, 0 if the parent is unsure whether the child shows the behavior or not, +1 when the child probably shows the behavior and +2 if the parent remembers a particular instance where the child exhibited the behavior. This is a retrospective tool - parents are asked to assess whether their child exhibited given behavior at any time in the past. Following the original version, the EMQ-PL scores are computed separately for the gross motor (GM), fine motor (FM), and perception-action integration (PA) domains and for the total score. There are 49 items in the GM scale (possible score range: from -98 to +98), 48 in the FM scale (possible score range: from -94 to +94) and 31 in the PA scale (possible score range: from -62 to +62; possible total score range: from -254 to +254).
Data collection
The EMQ-PL questionnaire and a demographics and health-related survey were completed by caregivers of infants and toddlers aged 2-25 months. Due to Covid-19 restrictions and national lockdown, the data collection was conducted entirely online. The caregivers were also asked to report whether their child was referred to a pediatric physiotherapist. If yes, they were asked to provide the reason for it. The adaptation was carried out as part of a larger research project, which obtained clearance from the Ethics Committee at the Institute of Psychology, Polish Academy of Sciences.
Data analysis
To investigate the psychometric properties of the EMQ-PL in the group of typically developing infants, we calculated Cronbach’s alpha. Pearson’s correlation coefficients were used to check the relations between the EMQ-PL subscale scores and the participant age. Partial correlations controlling for infant's age were used to check the relations between subscales and the total score of the EMQ-PL. To verify whether the EMQ-PL has equally good psychometric properties across all tested ages, we calculated Cronbach’s alpha separately for four age groups, representing every half-year (infants aged 1.9-6.0, 6.01-12.0, 12.01-18.0, 18.01-24.9 months). Data analysis was conducted in IBM SPSS Statistics 28; data was visualized using R, RStudio and ggplot2 package (Wickham, 2016).
To investigate differences in EMQ-PL scores between infants who were and were not referred for physiotherapy as well as potential sex differences across the wide age range, we have calculated age-independent scores (see Smith & Libertus, 2022 for detailed description and equations and Fig.1 for the distribution of scores in our sample). Next, we used independent t-tests, which are robust to large discrepancies in group size, to check the effects of sex and referral for physiotherapy on EMQ-PL age-independent scores.
Results
Descriptive Statistics
Descriptive statistics for EMQ-PL raw scores are presented in Table 2. Several participants in the two oldest groups reached maximum score in the GM subscale (first half of the second year: N = 1, second half of the second year: N = 3), similarly some participants from the oldest group reached max score for PA subscale (N = 9). In contrast, none of the infants reached max score in the FM subscale - the highest score was 84, which is 10 points below the max score. This suggests that items in the Fine Motor scale were the most difficult and differentiating in the Polish sample.
Table 2
Descriptive statistics of EMQ-PL scores
|
Age in months
|
N
|
Total Score
|
Gross Motor
|
Fine Motor
|
Perception/Action
|
Mean (SD)
|
Min
|
Max
|
Mean (SD)
|
Min
|
Max
|
Mean (SD)
|
Min
|
Max
|
Mean (SD)
|
Min
|
Max
|
Entire sample
|
640
|
31.16 (115.33)
|
-219
|
244
|
12.93 (52.28)
|
-86
|
98
|
0.15 (36.11)
|
-85
|
84
|
18.06 (29.65)
|
-54
|
62
|
1.9-6.0
|
92
|
-146.65 (29.78)
|
-219
|
-73
|
-61.98 (10.48)
|
-86
|
-28
|
-57.65 (13.06)
|
-85
|
-31
|
-24.02 (9.36)
|
-54
|
-10
|
6.01-12.0
|
217
|
-37.94 (53.90)
|
-138
|
99
|
-18.64 (27.18)
|
-67
|
59
|
-18.05 (16.70)
|
-58
|
27
|
-1.24 (14.68)
|
-24
|
47
|
12.01-18.0
|
195
|
96.11 (51.48)
|
-39
|
212
|
41.63 (29.25)
|
-35
|
98
|
19.09 (15.87)
|
-12
|
65
|
35.39 (13.85)
|
-6
|
59
|
18.01-25.0
|
136
|
166.41 (29.64)
|
72
|
244
|
72.94 (14.88)
|
-8
|
98
|
41.11 (15.18)
|
-1
|
84
|
52.46 (6,98)
|
28
|
62
|
Internal consistency
The Polish EMQ had excellent internal consistency for all subscales in all age groups (see Table 3). Cronbach’s alphas in all age groups were above 0.9 for the total score and above 0.7 for the three subscales (Cronbach’s alpha score of 1.0 is the max possible value). The lowest consistency was observed for GM in the first half of the first year of life (0.781) and for PA in the oldest group aged 18 to 25 months (0.783), which are still very high scores.
Table 3
Cronbach’s alphas for EMQ-PL in each age group
|
Age in months
|
N
|
Total score
|
GM
|
FM
|
PA
|
Entire sample
|
640
|
0.988
|
0.977
|
0.959
|
0.961
|
1.9-6.0
|
92
|
0.928
|
0.781
|
0.840
|
0.826
|
6.01-12.0
|
217
|
0.962
|
0.934
|
0.873
|
0.886
|
12.01-18.0
|
195
|
0.953
|
0.937
|
0.848
|
0.865
|
18.01-25.0
|
136
|
0.909
|
0.856
|
0.849
|
0.783
|
Correlations with age
The age was positively correlated with all EMQ-PL raw scores (see Figure 2): total score: r(638) = 0.946, p < .001; GM: r(638) = 0.921, p < 0.001; FM: r(638) = 0.925, p < .001; PA: r(638) = 0.928, p < .001.
Correlations between the subscales
We investigated if there were relations between the EMQ subscales. Partial correlations with age as a controlled variable were used to investigate the correlations between the subscales. All three subscales significantly correlated with each other and with the total score (Table 4). Interestingly, FM and PA were more correlated with each other than with GM score, which highlights the sensitivity of the EMQ to the relations between object manipulation and perceptual skills.
Table 4
Partial correlation coefficients between subscales and total score controlling for age
Variable
|
|
Total
|
GM
|
FM
|
GM
|
Pearson’s r
|
0.867
|
-
|
|
|
p-value
|
<.001
|
-
|
|
FM
|
Pearson’s r
|
0.821
|
0.507
|
-
|
|
p-value
|
<.001
|
<.001
|
-
|
PA
|
Pearson’s r
|
0.771
|
0.467
|
0.608
|
|
p-value
|
<.001
|
<.001
|
<.001
|
Sex differences: age-independent scores
To check if there were sex differences among the different subscales of the EMQ-PL, we used age-independent scores. The results did not show any sex differences in the Total score t(638) = -1.21, p = .227, or any of the subscales: GM: t(638) = -1.19, p = .236; FM: t(638) = -0.26, p = .795; PA: t(638) = -1.27, p = .203.
Pediatric physiotherapy: age-independent scores
To assess whether the EMQ-PL may be suitable as a screening tool, we compared the age-independent scores of infants who were and were not referred for pediatric physiotherapy. Infants who were referred for pediatric physiotherapy had significantly lower GM scores (N = 202, M = 47.58, SD = 3.95) compared to those who did not (N = 438, M = 49.31, SD = 3.45), two-sided t-test, t(638) = 5.63, p < .001, Cohen’s d = 0.48 (see Figure 3). Similarly, infants who were referred for physiotherapy had lower Total score t(638) = 3.44, p < .001, Cohen’s d = 0.29. No differences were found for the other two subscales FM: t(638) = 0.45, p = .652; PA: t(638) = 0.01, p = .989.
Discussion
Here, we presented the process of adaptation of the Early Motor Questionnaire to Polish language (EMQ-PL) and its validation in a large sample (N = 640) of Polish infants aged 2 to 24 months. We showed that EMQ has excellent internal consistency across all measured ages and its raw scores in all subscales are highly correlated with infants’ age. Furthermore, using age-independent scores (Smith & Libertus, 2022) we found that the EMQ-PL was sensitive at the group level to differentiate between infants who were and were not referred for physiotherapy. Furthermore, in our sample we have not found any sex-related differences in any of the subscales.
In the online study we were not able to investigate the concurrent validity of any of the EMQ-PL subscales. For this reason, we recruited an independent sample of Polish infants for a lab-based study and examined the concurrent validity of the EMQ-PL Gross Motor and Total scores and the experimenter-administered Alberta Infant Motor Scale (AIMS, Piper & Darrah, 1994) at four time points across the first year of life, when infants were around 4, 6, 9 and 12 months of age. We selected these ages as they reflect significant changes in motor control and gross motor development.