Are Adolescents with a Wider Vocabulary Faster at Inference Making During Reading? Evidence from Self-Paced Reading
John Kirby
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
I have indicated "no answer" above because I think that further description and analyses are required.
This paper assesses the effect of vocabulary knowledge on inference making in 265 Spanish-speaking adolescents, ranging in age from 11 to 18. The main findings are that having to make an inference, instead of reading a word repeated from a previous sentence slows reading, and that this effect is greater for adolescents with lower vocabulary scores.
The paper is well-written (a few problems are noted below), and the analyses are carried out and reported appropriately. The research appears to address a gap in the literature, in that both younger and older participants have been studied previously.
There are a few problems with how the Methods are described and the data analyzed. If these can be resolved successfully, I think the paper would make a valuable contribution to the literature.
Abstract
It should be stated that the participants are Spanish-speaking and where they were recruited.
The large age range reported here immediately raises the question whether age is controlled.
Introduction
This section is good, but several questions occurred to me as I read it; the authors should consider addressing these.
- Is the effect of vocabulary knowledge specific to the words involved in the inference, or is it more of a general (ability) effect? Would teaching the words involved in the inference improve performance?
- What is the nature of the inferences that the readers in these studies have to make? This issue comes up again in the Method.
Method
Page 5, line 199 Indicate where and how the participants were recruited.
5, 200 What is the SD of age?
5, line 208, Table 1 English translations of the Spanish sentences would help.
What type or types of inferences do the participants have to make? For instance, are they always supplying a missing word, or are there inferences about what a pronoun or synonym means? Are they relatively shallow or do some require deeper inferencing? The authors need to describe the inferences much more completely. If there is variety in their types, some sort of tabulation would help. (Later I will ask if the type of inference, or some other characteristic, contributes to reading time.)
What was the reading level of the sentences that the participants had to read? This is especially important regarding the words involved in the inferences.
Regarding the Procedure, is there any check that the participants are actually reading the sentences? What is to stop some, perhaps some with low vocabulary scores, from just moving on? Clearly most are reading for meaning, because the overall pattern of the results varies by condition, but there are some questions about the low-vocabulary students performed. Is it possible that the reading time cut-off scores need to be changed?
Vocabulary measure. This test will be unfamiliar to most of the journal’s readers, so it needs to be described as fully as possible. Relevant psychometric data on reliability and validity should be supplied. I would note that as a receptive vocabulary measure it is probably more one of breadth than depth.
Data analysis
When I first read this, I wondered if using only the reading time of the first word was enough of a control. The authors may wish to explain how this choice was made. Why not take the time of reading more text?
5, 230 should be mixed-effects
5, 234 The reason for examining only the first six words should be explained more completely here.
Results
The major problem in the results is that no account of age is taken. We know that vocabulary increases with age, perhaps especially in the adolescent period. Therefore the vocabulary effects that are so impressive in Figures 1, 2, and 3 are confounded with age. Age should be covaried in all of the analyses.
The age problem may explain some of the variability seen in Figures 2 and 3. Some of the low vocabulary participants show very fast reading times. Figure 3 shows many participants with negative inference costs in the low vocabulary range. How can this be?
The authors compare inference to repetition conditions, but do not examine variability due to different inferences or types of inferences. It would be interesting to know if the inferences with the highest reading times had particular natures, syntactic structures, or difficult vocabulary.
7, 278 The authors state the effect of the complete range of vocabulary scores (35-91) on inference times. These are of course extreme scores; it would be more informative to show the effect over a more modest range, for instance +/- 1 SD.
Figure 3 is impressive, but it gives the false impression (which the authors do not state) that vocabulary scores over 90 would show negligible inference costs. That would only be true for the inferences used here. I realize the authors do not state this, but it might be be helpful to be clear that these results are only for these particular inferences.
Discussion
Some comment on inference difficulty should be made. This could be both in terms of their Inference Cost variable and relative to vocabulary level of the text, and other variables concerning the inferences.
9, 325 The comment about “deeper” vocabulary knowledge raises questions about the authors’ vocabulary measure. As a receptive measure, it is probably more a measure of vocabulary breadth than vocabulary depth. Investigating the effect of vocabulary depth on inferencing could be suggested as a topic for future research. The authors’ results relate to more “extensive” vocabulary knowledge, not deeper.
Minor corrections
Page 2, line 68 play should be plays
2, 77 “younger” this wording says the authors want research with readers younger than children. Please revise.
3, 136 “did explored” should be “explored”
4, 151 “revised” should be “reviewed”
Author Response
Dear Reviewer 1, below you will find a detailed response to all your comments, and those to Reviewer 2. I opted for provinding responses to both of you, since there are many similar points, but some contractory as well. In the revised manuscript, you will find all changes marked now in blue font.
Response to Reviewers
Reviewer 1
This paper assesses the effect of vocabulary knowledge on inference making in 265 Spanish-speaking adolescents, ranging in age from 11 to 18. The main findings are that having to make an inference, instead of reading a word repeated from a previous sentence slows reading, and that this effect is greater for adolescents with lower vocabulary scores.
The paper is well-written (a few problems are noted below), and the analyses are carried out and reported appropriately. The research appears to address a gap in the literature, in that both younger and older participants have been studied previously.
There are a few problems with how the Methods are described and the data analyzed. If these can be resolved successfully, I think the paper would make a valuable contribution to the literature.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Abstract
It should be stated that the participants are Spanish-speaking and where they were recruited.
Response: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The large age range reported here immediately raises the question whether age is controlled.
Response: The new analysis includes age as a predictor in the model. The results remain.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Introduction
This section is good, but several questions occurred to me as I read it; the authors should consider addressing these.
- Is the effect of vocabulary knowledge specific to the words involved in the inference, or is it more of a general (ability) effect? Would teaching the words involved in the inference improve performance?
Response: The effect of vocabulary knowledge in inferencing appears to be more indicative of a general cognitive ability rather than one specifically tied to the words involved. The words we used in the experimental materials were common and familiar to most readers, suggesting that inferencing does not hinge on isolated word recognition. Instead, a larger vocabulary contributes to the construction of a more robust situational model (we believe this is also true for deeper vocabulary, but in our study, we only have a measure of vocabulary breadth). This enhanced situational model provides the representational context that allows readers to process and generate inferences efficiently, with less cognitive effort. As a result, vocabulary knowledge facilitates inferencing by enabling a coherent and accessible situational model, rather than through specific word recognition alone. In other words, teaching any vocabulary would enhance inference making during reading.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- What is the nature of the inferences that the readers in these studies have to make? This issue comes up again in the Method.
Response: We apologize for the misspecification of the type of inferences we are evaluating. They are bridging inferences (see Graesser et al., 1994; Singer & Halldorson, 1996). We have now added a concrete description of the type of inference we assessed in the introduction.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Method
Page 5, line 199 Indicate where and how the participants were recruited.
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, 200 What is the SD of age?
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, line 208, Table 1 English translations of the Spanish sentences would help.
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
What type or types of inferences do the participants have to make? For instance, are they always supplying a missing word, or are there inferences about what a pronoun or synonym means? Are they relatively shallow or do some require deeper inferencing? The authors need to describe the inferences much more completely. If there is variety in their types, some sort of tabulation would help. (Later I will ask if the type of inference, or some other characteristic, contributes to reading time.)
Response: They are all bridging inferences (Singer et al., 1992, see also Graesser et al., 1994; Singer & Halldorson, 1996). All items are constructed the same way. We have now added a concrete description of the type of inference we assessed in the introduction.
Graesser, A. C., Singer, M., & Trabasso, T. (1994). Constructing inferences during narrative text comprehension. Psychological Review, 101(3), 371–95.
Singer, M., & Halldorson, M. (1996). Constructing and validating motive bridging inferences. Cognitive Psychology, 30(1), 1–38.
Singer, M., Andruslak, P., Reisdorf, P., & Black, N. L. (1992). Individual differences in bridging inference processes. Memory & Cognition, 20(5), 539-548.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
What was the reading level of the sentences that the participants had to read? This is especially important regarding the words involved in the inferences.
Response: Sentences are all very simple sentences with common and frequent words. All materials are available at https://osf.io/ykfg9/?view_only=0025e1d126c943c49f5f216b482cdcef.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Regarding the Procedure, is there any check that the participants are actually reading the sentences? What is to stop some, perhaps some with low vocabulary scores, from just moving on? Clearly most are reading for meaning, because the overall pattern of the results varies by condition, but there are some questions about the low-vocabulary students performed. Is it possible that the reading time cut-off scores need to be changed?
Response: In this self-paced reading study, participants control the progression through the text by pressing a button, which prevents them from disengaging, as the experiment will not proceed unless they actively signal comprehension by moving forward. Additionally, to ensure participants were paying attention and processing the material, we included occasional comprehension questions. These questions served as attention-gatherer to confirm that participants were engaging with the text meaningfully. This well-established approach design minimizes the likelihood of participants bypassing content without understanding.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Vocabulary measure. This test will be unfamiliar to most of the journal’s readers, so it needs to be described as fully as possible. Relevant psychometric data on reliability and validity should be supplied. I would note that as a receptive vocabulary measure it is probably more one of breadth than depth.
Response: We have provided a better description of the test.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Data analysis
When I first read this, I wondered if using only the reading time of the first word was enough of a control. The authors may wish to explain how this choice was made. Why not take the time of reading more text?
5, 230 should be mixed-effects
Response: Corrected
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, 234 The reason for examining only the first six words should be explained more completely here.
Response: Added. We did this to provide an overview of the extent of effect of condition.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Results
The major problem in the results is that no account of age is taken. We know that vocabulary increases with age, perhaps especially in the adolescent period. Therefore the vocabulary effects that are so impressive in Figures 1, 2, and 3 are confounded with age. Age should be covaried in all of the analyses.
Response: We have added age to the model and the critical results, i.e., the interaction between condition and vocabulary, remains.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The age problem may explain some of the variability seen in Figures 2 and 3. Some of the low vocabulary participants show very fast reading times. Figure 3 shows many participants with negative inference costs in the low vocabulary range. How can this be?
Response: Indeed, some participants exhibit longer reading times in the repetition condition compared to the inference condition. This variation likely reflects noise inherent in experimental data, not uncommon at all in online reading measures. Such fluctuations are expected in online processing research and do not undermine the main findings, as these cases are distributed across the full range of vocabulary scores rather than clustering at specific points. Thus, while individual errors may introduce some variability, this does not systematically impact the overall relationship between vocabulary and inference cost observed in our study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The authors compare inference to repetition conditions, but do not examine variability due to different inferences or types of inferences. It would be interesting to know if the inferences with the highest reading times had particular natures, syntactic structures, or difficult vocabulary.
Response: This is because there are no different types of inference in the study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
7, 278 The authors state the effect of the complete range of vocabulary scores (35-91) on inference times. These are of course extreme scores; it would be more informative to show the effect over a more modest range, for instance +/- 1 SD.
Response: If we calculate the mean of vocabulary by participants this is equal to 63.31 and the SD is equal to 10.46. Thus, +/- 1 SD around the mean gives a range of 20.92 (min=52.85, max= 73.77). Now, if we calculate the change per 10 unit of vocabulary is equal to 4.46 millisecond, and thus, the change in reading times for a vocabulary range of +/- 1 SD around the mean is equal to 9.34 milliseconds, of 14.01 for +/- 1.5 SD, and 18.68 for +/- 2 SD. We have now added this information to the manuscript.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Figure 3 is impressive, but it gives the false impression (which the authors do not state) that vocabulary scores over 90 would show negligible inference costs. That would only be true for the inferences used here. I realize the authors do not state this, but it might be be helpful to be clear that these results are only for these particular inferences.
Response: Thanks for the suggestion. We have added this information in the discussion section as part of the limitations of the study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Discussion
Some comment on inference difficulty should be made. This could be both in terms of their Inference Cost variable and relative to vocabulary level of the text, and other variables concerning the inferences.
Response: It is difficult to estimate the difficulty of each inference, but we aimed to have relatively easy inferences. All materials are available at https://osf.io/ykfg9/?view_only=0025e1d126c943c49f5f216b482cdcef
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
9, 325 The comment about “deeper” vocabulary knowledge raises questions about the authors’ vocabulary measure. As a receptive measure, it is probably more a measure of vocabulary breadth than vocabulary depth. Investigating the effect of vocabulary depth on inferencing could be suggested as a topic for future research. The authors’ results relate to more “extensive” vocabulary knowledge, not deeper.
Response: While it is true that the vocabulary measure pertains to the breadth of vocabulary dimension, the hypothesis is still open to both dimensions, that is breadth and depth. It is important to notice that a preliminary study we cite used a very similar experimental paradigm in adults, and their vocabulary measure was that of the subtest of vocabulary from the Wechler inventory, and thus, measure vocabulary depth. They also found that better vocabulary was associated with lower inferential processing costs as reflected in reading times. Considering this as part of the understanding we have about the relation between vocabulary and inference making it seems unfair to limit our hypothesis to vocabulary breadth.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Minor corrections
Page 2, line 68 play should be plays
2, 77 “younger” this wording says the authors want research with readers younger than children. Please revise.
3, 136 “did explored” should be “explored”
4, 151 “revised” should be “reviewed”
Response: All corrected. We thank the reviewer for the thorough review of our manuscript.
Reviewer 2
This article sets out to establish whether adolescents with a larger receptive vocabulary process overt and non-overt (inferred) referents at greater speed than adolescents with a smaller receptive vocabulary. After a presentation of a range of previous studies examining the effects of the breadth and depth of vocabulary on comprehension and processing speed in children and adults, this study provides empirical data showing a clear relation between vocabulary knowledge and processing speed in Spanish adolescents.
The introduction, consisting mostly of a (somewhat unordered) literature review, is a little long (146 lines) in comparison with the rest of the article (187 lines), which includes methodology, data and discussion.
The description of the experiment’s setup is rather brief, but links to additional data are provided. The data is generally clearly presented and illustrated in appropriate graphs; the discussion is generally clear and to the point.
However, the analysis of the data has a logical weakness. The author appears to take it for granted that the relation between receptive vocabulary and reading/processing speed is a causal one, but that is not necessarily so; we may well be dealing with a spurious correlation, in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Specific comments and suggested changes
Title: I strongly suggest changing the title to make it clearer, more comprehensible and more attractive to potential readers. The current title is an accurate summary of the findings of the article, but it is quite dense, which makes it difficult to process. Rather than saying that “the vocabulary modulates speed of word integration”, something more general could attract a wider readership, perhaps something along the lines of “Are adolescents with a wider vocabulary quicker at resolving lexical inferences? Evidence from self-paced reading”
Response: What about “Are adolescents with a wider vocabulary faster at inference making during reading? Evidence from self-paced reading”
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Introduction/literature review: As mentioned in the previous section, the introduction consists almost entirely of an overview of related studies. This literature overview lacks order and is, at times, a little repetitive. Thus, in lines 75-82, lines 128-135 and lines 162-165 the author repeatedly observes that there is a lack of studies of adolescents in this area, before going on to summarise yet more existing studies of certain age groups. This section would benefit from being restructured and cleaned up.
Response: We have carefully examined the sections the reviewer pointed out. The first text segment (lines 75-82) offers a general comment on the relation between vocabulary and text comprehension, and the lack of studies in our population of interest. The second segment (lines 128-135) is more specific, referring to the relation between vocabulary and inference making. Only the last segment (lines 162-165) is, in our view, a bit repetitive and thus we have deleted it.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Vocabulary “depth”/ “richness”: There is a brief explanation of what is meant by this (l.38-40), but it is not very clear how this variable can be quantified or whether it is clearly defined and comparable in different studies. As “vocabulary depth” (defined as “the richness of understanding about those words, including semantic associations, morphological characteristics, and syntactic roles”, l.38-40) does not appear to be a variable used in the study presented in this article, that’s not a major problem.
However, throughout the description and discussion of the study, the terms “richer vocabularies” (l.18, 299, 319, 336), “richer vocabulary knowledge” (l. 172, 193, 287) are used, though presumably to refer to the number of words known by participants (i.e. “vocabulary breadth”). This terminological issue should be sorted out.
Response: We thank the reviewer for the detailed scrutiny of the manuscript. Indeed, “richer” was misused in those places the reviewer pointed out. We have corrected all those instances now, changing the phrase “richer vocabularies” for “larger vocabularies” and “richer vocabulary knowledge” for “wider vocabulary knowledge”.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
“Methods” and “Data analysis” sections: Information such as the number of participants is not really part of the method, but simply of the experiment’s setup. However, more information about the participants is needed. In which country was the experiment conducted? (There are native Spanish speakers in many countries.) How were the informants recruited? Did they sign informed consent forms?
Response: The revised manuscript includes this information.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
“Vocabulary” section: This section, consisting of a single sentence, is insufficient. The vocabulary range of the participants in this experiment is the fundamental variable in this study, and the reader should be told how the test works, what kind of lexical items the participants are asked to identify, how many words TEVI-R tests for, etc.; a single reference to Echeverría, Herrera & Segure (2015) is insufficient. More importantly still, the author should explain why this test is the best way of measuring participants’ vocabulary range for this study. It’s a test “standardized for the Chilean population”, but there is no indication in which country the experiment was conducted. TEVI-R can be used for children from the age of 30 months; wouldn’t it be better to use a test tailored specifically for adolescents in order to better determine their vocabulary range?
Response: This section now includes further information about the test and its nature.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Data analysis: This section clearly describes a crucial part of the methodology used in this study, so it should really be part of the Methodology section.
It is not very clear which time intervals the “reading times” refer to. In line 227 it says “we used the reading time TO the first word in the target sentence as control”, but it is not clear what the starting point of the time interval is. Is it the time from the beginning of the context sentence to the first word of the target sentence? In line 228 it says “the reading time FOR the first word”, which doesn’t help to clarify things. In lines 293-295, the author refers to “reading time FOR the word preceding the critical word” (presumably the article), but also “reading times AT the critical word”. These different prepositions are extremely confusing, and it is absolutely essential that the author clarify the exact time intervals that were measured. Is it simply the time it took the respective participant to read the control word/ the critical word?
Response: We apologize for the misuse of the prepositions, which we have corrected in the revised version of the manuscript. Indeed, as a self-paced reading study, any reading time, corresponds to a give word, and therefore we should say “reading times for the target word”, for instance.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
In any case, it seems odd to use the first word in the target sentence as control for the individual’s reading speed, as the first word of the target sentence can be very short, e.g. a definite article, as seen in the example in Table 1, which is unlikely to reveal useful information about the individual’s reading speed because it can be recognized at first sight and doesn’t involve sequential reading as such. Why wasn’t the control reading time measured for a word in the context sentence, or the entire context sentence, where effects of inference or repetition can be ruled out with more confidence?
Response: There were several key reasons for selecting the first word of the target sentence in our experimental materials as the control for reading speed. First, we focused our analysis solely on the data from the target sentence because its structure and content are constant across experimental conditions, which helps to isolate the effects of vocabulary and inference. Second, this initial word showed no significant difference in reading time between conditions, making it a stable baseline for determining overall reading speed. Third, because it immediately precedes the critical word, it serves as the closest temporal baseline just before the experimental manipulation comes into effect. This proximity allows us to control for individual reading speed immediately before the condition effects “kick in.” While it might seem that a short word, such as a definite article, offers limited information about reading speed, its placement provides a reliable measure of transition into the target sentence, capturing individual differences without confounding effects from the varied processing demands in the context sentence. Using a control word from the context sentence, or the entire context sentence, would have introduced additional variability due to the differences in length (between conditions), potentially impacting our baseline reading speed measure.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Table 1/Mixed effects: The data in Table 1 needs a little more explanation, in particular the values given in the last line, “Vocabulary x Condition”. Though the p-value does indicate statistical significance if you take p<0.05 as your threshold, there is little clarity and no discussion about the choice of the threshold(s) for the p-values. This discussion should be added. Can any conclusions be drawn from the differences between the p-values for different variables?
Response: We adopted a p-value threshold of p < 0.05 to determine statistical significance, following conventional practices in cognitive and educational research, and psychology more generally. The significance of the "Vocabulary x Condition" interaction (p = 0.023) suggests a meaningful interaction effect, indicating that the impact of vocabulary on reading times differs across conditions (Repetition vs. Inference). Since interaction between discrete (condition) and continuous (vocabulary) variables are difficult to interpret based only on the output of the regression model we provide two further graphs (Figures 2 and 3). As it can be seen in those plots, the interaction effects implies that vocabulary knowledge more strongly influences reading efficiency when an inference is required, as shown by steeper slopes in the Inference condition. Importantly, this is true for the critical word and not for the baseline.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
References: It is surprising that the author does not mention the following study, which is very similar in scope to the one presented in this article:
Pezoa, José Pablo Pezoa & Pelusa Orellana. 2021. “La relación entre comprensión lectora y vocabulario receptivo en estudiantes chilenos: un estudio exploratorio”, OCNOS Revista De Estudios Sobre Lectura 20 (2), 7-20.
Response: We were aware of the cited work, which is quite different from our study. First, it focused on 5th-graders, just like other studies we cited, not in adolescents. Second, it does not explore the relation between vocabulary and inference making, as we do, but rather on the relation between vocabulary and general reading comprehension. Third, it does not provide any real-time processing measure, such as self-paced reading (as we do), or eye tracking reading. Finally, it is exploratory in nature, and thus, it does not test any particular hypothesis, unlike our study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Comments on the Quality of English Language
The English language is generally clear and fluent, but with a number of grammatical, lexical and stylistic errors, some of which are listed below. The revised version of this article should be gone over by a native speaker if possible.
l.48, 51: “This means... This suggests...” Style: avoid repetition
l.54: “how vocabulary knowledge influences the cognitive cost” > “how vocabulary knowledge influences cognitive cost”
l.68: “working memory also play a role” > “working memory also plays a role”
l.72: “Other have also found that...” > “Others have also found that...”
l.85: “understand” > “explain”
l.93: “Their findings showed that inferences based on distant information, were significantly more challenging”— Remove comma before “were”.
l.100/101: “this early study... focus on” > “this early study... focused on”
l.136: “In this sense” is a literal translation of the Spanish linker “En este sentido” that cannot be used in the same way in English, as no “sense” has previously been mentioned.
l.136: “did explored” > “did explore” or “explored”
l.150: “underscoring again the pivotal role” > “underscoring, again, the pivotal role” or “again underscoring the pivotal role”
l.200: The mean age (M = 16.2). > The participants’ mean age was just over 16 years (M = 16.2).
l.200-201: There is no need to spell out large numbers. “153 were female, 111 male, and one...”
- 206, 240: “can be found in” > “can be found at”
l.204, 211: “short stories”, “stories” The author should find a different term to refer to these sequences of two sentences; they’re not stories.
l.215: “a critical word positioned always as” > “a critical word always positioned as”
- 216: “we measure” > “we measured”
l.228: “we submitted reading times” Choose a more appropriate verb.
l.231-232: “times below the lower 1% and above the 99%” Rephrase, perhaps “times below 1% and above 99%”
l.233: “we plot” > “we plotted”
l.236: “we compute” > “we computed”
l.239: Full stop after “subject” missing.
l.247: Check the punctuation; one of the periods should probably be a comma.
l.248: “spilled over the three words following the critical” > “spilled over to the three words following the critical one”
l.283: “the study was aimed to investigate” > “the study aimed to investigate”
l.317-318: “vocabulary’s impact” > “the impact of vocabulary”
Response: We profoundly appreciate the work and commitment of the reviewer in proof-reading our manuscript to such depth. We have corrected all errors pointed by the reviewer.
Reviewer 2 Report
Comments and Suggestions for Authors
Review of the article “Adolescents’ vocabulary modulates speed of word integration into preceding text across sentence boundaries: Evidence from self-paced reading”, submitted for publication in the MDPI journal Language Sciences in October 2024.
General
This article sets out to establish whether adolescents with a larger receptive vocabulary process overt and non-overt (inferred) referents at greater speed than adolescents with a smaller receptive vocabulary. After a presentation of a range of previous studies examining the effects of the breadth and depth of vocabulary on comprehension and processing speed in children and adults, this study provides empirical data showing a clear relation between vocabulary knowledge and processing speed in Spanish adolescents.
The introduction, consisting mostly of a (somewhat unordered) literature review, is a little long (146 lines) in comparison with the rest of the article (187 lines), which includes methodology, data and discussion.
The description of the experiment’s setup is rather brief, but links to additional data are provided. The data is generally clearly presented and illustrated in appropriate graphs; the discussion is generally clear and to the point.
However, the analysis of the data has a logical weakness. The author appears to take it for granted that the relation between receptive vocabulary and reading/processing speed is a causal one, but that is not necessarily so; we may well be dealing with a spurious correlation, in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
Specific comments and suggested changes
Title: I strongly suggest changing the title to make it clearer, more comprehensible and more attractive to potential readers. The current title is an accurate summary of the findings of the article, but it is quite dense, which makes it difficult to process. Rather than saying that “the vocabulary modulates speed of word integration”, something more general could attract a wider readership, perhaps something along the lines of “Are adolescents with a wider vocabulary quicker at resolving lexical inferences? Evidence from self-paced reading”
Introduction/literature review: As mentioned in the previous section, the introduction consists almost entirely of an overview of related studies. This literature overview lacks order and is, at times, a little repetitive. Thus, in lines 75-82, lines 128-135 and lines 162-165 the author repeatedly observes that there is a lack of studies of adolescents in this area, before going on to summarise yet more existing studies of certain age groups. This section would benefit from being restructured and cleaned up.
Vocabulary “depth”/ “richness”: There is a brief explanation of what is meant by this (l.38-40), but it is not very clear how this variable can be quantified or whether it is clearly defined and comparable in different studies. As “vocabulary depth” (defined as “the richness of understanding about those words, including semantic associations, morphological characteristics, and syntactic roles”, l.38-40) does not appear to be a variable used in the study presented in this article, that’s not a major problem. However, throughout the description and discussion of the study, the terms “richer vocabularies” (l.18, 299, 319, 336), “richer vocabulary knowledge” (l. 172, 193, 287) are used, though presumably to refer to the number of words known by participants (i.e. “vocabulary breadth”). This terminological issue should be sorted out.
“Methods” and “Data analysis” sections: Information such as the number of participants is not really part of the method, but simply of the experiment’s setup. However, more information about the participants is needed. In which country was the experiment conducted? (There are native Spanish speakers in many countries.) How were the informants recruited? Did they sign informed consent forms?
“Vocabulary” section: This section, consisting of a single sentence, is insufficient. The vocabulary range of the participants in this experiment is the fundamental variable in this study, and the reader should be told how the test works, what kind of lexical items the participants are asked to identify, how many words TEVI-R tests for, etc.; a single reference to Echeverría, Herrera & Segure (2015) is insufficient. More importantly still, the author should explain why this test is the best way of measuring participants’ vocabulary range for this study. It’s a test “standardized for the Chilean population”, but there is no indication in which country the experiment was conducted. TEVI-R can be used for children from the age of 30 months; wouldn’t it be better to use a test tailored specifically for adolescents in order to better determine their vocabulary range?
Data analysis: This section clearly describes a crucial part of the methodology used in this study, so it should really be part of the Methodology section.
It is not very clear which time intervals the “reading times” refer to. In line 227 it says “we used the reading time TO the first word in the target sentence as control”, but it is not clear what the starting point of the time interval is. Is it the time from the beginning of the context sentence to the first word of the target sentence? In line 228 it says “the reading time FOR the first word”, which doesn’t help to clarify things. In lines 293-295, the author refers to “reading time FOR the word preceding the critical word” (presumably the article), but also “reading times AT the critical word”. These different prepositions are extremely confusing, and it is absolutely essential that the author clarify the exact time intervals that were measured. Is it simply the time it took the respective participant to read the control word/ the critical word?
In any case, it seems odd to use the first word in the target sentence as control for the individual’s reading speed, as the first word of the target sentence can be very short, e.g. a definite article, as seen in the example in Table 1, which is unlikely to reveal useful information about the individual’s reading speed because it can be recognized at first sight and doesn’t involve sequential reading as such. Why wasn’t the control reading time measured for a word in the context sentence, or the entire context sentence, where effects of inference or repetition can be ruled out with more confidence?
Table 1/Mixed effects: The data in Table 1 needs a little more explanation, in particular the values given in the last line, “Vocabulary x Condition”. Though the p-value does indicate statistical significance if you take p<0.05 as your threshold, there is little clarity and no discussion about the choice of the threshold(s) for the p-values. This discussion should be added. Can any conclusions be drawn from the differences between the p-values for different variables?
References: It is surprising that the author does not mention the following study, which is very similar in scope to the one presented in this article:
Pezoa, José Pablo Pezoa & Pelusa Orellana. 2021. “La relación entre comprensión lectora y vocabulario receptivo en estudiantes chilenos: un estudio exploratorio”, OCNOS Revista De Estudios Sobre Lectura 20 (2), 7-20.
Comments for author File: 
Comments.pdf
Comments on the Quality of English Language
The English language is generally clear and fluent, but with a number of grammatical, lexical and stylistic errors, some of which are listed below. The revised version of this article should be gone over by a native speaker if possible.
l.48, 51: “This means... This suggests...” Style: avoid repetition
l.54: “how vocabulary knowledge influences the cognitive cost” > “how vocabulary knowledge influences cognitive cost”
l.68: “working memory also play a role” > “working memory also plays a role”
l.72: “Other have also found that...” > “Others have also found that...”
l.85: “understand” > “explain”
l.93: “Their findings showed that inferences based on distant information, were significantly more challenging”— Remove comma before “were”.
l.100/101: “this early study... focus on” > “this early study... focused on”
l.136: “In this sense” is a literal translation of the Spanish linker “En este sentido” that cannot be used in the same way in English, as no “sense” has previously been mentioned.
l.136: “did explored” > “did explore” or “explored”
l.150: “underscoring again the pivotal role” > “underscoring, again, the pivotal role” or “again underscoring the pivotal role”
l.200: The mean age (M = 16.2). > The participants’ mean age was just over 16 years (M = 16.2).
l.200-201: There is no need to spell out large numbers. “153 were female, 111 male, and one...”
l. 206, 240: “can be found in” > “can be found at”
l.204, 211: “short stories”, “stories” The author should find a different term to refer to these sequences of two sentences; they’re not stories.
l.215: “a critical word positioned always as” > “a critical word always positioned as”
l. 216: “we measure” > “we measured”
l.228: “we submitted reading times” Choose a more appropriate verb.
l.231-232: “times below the lower 1% and above the 99%” Rephrase, perhaps “times below 1% and above 99%”
l.233: “we plot” > “we plotted”
l.236: “we compute” > “we computed”
l.239: Full stop after “subject” missing.
l.247: Check the punctuation; one of the periods should probably be a comma.
l.248: “spilled over the three words following the critical” > “spilled over to the three words following the critical one”
l.283: “the study was aimed to investigate” > “the study aimed to investigate”
l.317-318: “vocabulary’s impact” > “the impact of vocabulary”
Author Response
Dear Reviewer 2, below you will find a detailed response to all your comments, and those to Reviewer 1. I opted for provinding responses to both of you, since there are many similar points, but some contractory as well. In the revised manuscript, you will find all changes marked now in blue font.
Response to Reviewers
Reviewer 1
This paper assesses the effect of vocabulary knowledge on inference making in 265 Spanish-speaking adolescents, ranging in age from 11 to 18. The main findings are that having to make an inference, instead of reading a word repeated from a previous sentence slows reading, and that this effect is greater for adolescents with lower vocabulary scores.
The paper is well-written (a few problems are noted below), and the analyses are carried out and reported appropriately. The research appears to address a gap in the literature, in that both younger and older participants have been studied previously.
There are a few problems with how the Methods are described and the data analyzed. If these can be resolved successfully, I think the paper would make a valuable contribution to the literature.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Abstract
It should be stated that the participants are Spanish-speaking and where they were recruited.
Response: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The large age range reported here immediately raises the question whether age is controlled.
Response: The new analysis includes age as a predictor in the model. The results remain.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Introduction
This section is good, but several questions occurred to me as I read it; the authors should consider addressing these.
- Is the effect of vocabulary knowledge specific to the words involved in the inference, or is it more of a general (ability) effect? Would teaching the words involved in the inference improve performance?
Response: The effect of vocabulary knowledge in inferencing appears to be more indicative of a general cognitive ability rather than one specifically tied to the words involved. The words we used in the experimental materials were common and familiar to most readers, suggesting that inferencing does not hinge on isolated word recognition. Instead, a larger vocabulary contributes to the construction of a more robust situational model (we believe this is also true for deeper vocabulary, but in our study, we only have a measure of vocabulary breadth). This enhanced situational model provides the representational context that allows readers to process and generate inferences efficiently, with less cognitive effort. As a result, vocabulary knowledge facilitates inferencing by enabling a coherent and accessible situational model, rather than through specific word recognition alone. In other words, teaching any vocabulary would enhance inference making during reading.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- What is the nature of the inferences that the readers in these studies have to make? This issue comes up again in the Method.
Response: We apologize for the misspecification of the type of inferences we are evaluating. They are bridging inferences (see Graesser et al., 1994; Singer & Halldorson, 1996). We have now added a concrete description of the type of inference we assessed in the introduction.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Method
Page 5, line 199 Indicate where and how the participants were recruited.
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, 200 What is the SD of age?
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, line 208, Table 1 English translations of the Spanish sentences would help.
RESPONSE: Added
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
What type or types of inferences do the participants have to make? For instance, are they always supplying a missing word, or are there inferences about what a pronoun or synonym means? Are they relatively shallow or do some require deeper inferencing? The authors need to describe the inferences much more completely. If there is variety in their types, some sort of tabulation would help. (Later I will ask if the type of inference, or some other characteristic, contributes to reading time.)
Response: They are all bridging inferences (Singer et al., 1992, see also Graesser et al., 1994; Singer & Halldorson, 1996). All items are constructed the same way. We have now added a concrete description of the type of inference we assessed in the introduction.
Graesser, A. C., Singer, M., & Trabasso, T. (1994). Constructing inferences during narrative text comprehension. Psychological Review, 101(3), 371–95.
Singer, M., & Halldorson, M. (1996). Constructing and validating motive bridging inferences. Cognitive Psychology, 30(1), 1–38.
Singer, M., Andruslak, P., Reisdorf, P., & Black, N. L. (1992). Individual differences in bridging inference processes. Memory & Cognition, 20(5), 539-548.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
What was the reading level of the sentences that the participants had to read? This is especially important regarding the words involved in the inferences.
Response: Sentences are all very simple sentences with common and frequent words. All materials are available at https://osf.io/ykfg9/?view_only=0025e1d126c943c49f5f216b482cdcef.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Regarding the Procedure, is there any check that the participants are actually reading the sentences? What is to stop some, perhaps some with low vocabulary scores, from just moving on? Clearly most are reading for meaning, because the overall pattern of the results varies by condition, but there are some questions about the low-vocabulary students performed. Is it possible that the reading time cut-off scores need to be changed?
Response: In this self-paced reading study, participants control the progression through the text by pressing a button, which prevents them from disengaging, as the experiment will not proceed unless they actively signal comprehension by moving forward. Additionally, to ensure participants were paying attention and processing the material, we included occasional comprehension questions. These questions served as attention-gatherer to confirm that participants were engaging with the text meaningfully. This well-established approach design minimizes the likelihood of participants bypassing content without understanding.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Vocabulary measure. This test will be unfamiliar to most of the journal’s readers, so it needs to be described as fully as possible. Relevant psychometric data on reliability and validity should be supplied. I would note that as a receptive vocabulary measure it is probably more one of breadth than depth.
Response: We have provided a better description of the test.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Data analysis
When I first read this, I wondered if using only the reading time of the first word was enough of a control. The authors may wish to explain how this choice was made. Why not take the time of reading more text?
5, 230 should be mixed-effects
Response: Corrected
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
5, 234 The reason for examining only the first six words should be explained more completely here.
Response: Added. We did this to provide an overview of the extent of effect of condition.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Results
The major problem in the results is that no account of age is taken. We know that vocabulary increases with age, perhaps especially in the adolescent period. Therefore the vocabulary effects that are so impressive in Figures 1, 2, and 3 are confounded with age. Age should be covaried in all of the analyses.
Response: We have added age to the model and the critical results, i.e., the interaction between condition and vocabulary, remains.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The age problem may explain some of the variability seen in Figures 2 and 3. Some of the low vocabulary participants show very fast reading times. Figure 3 shows many participants with negative inference costs in the low vocabulary range. How can this be?
Response: Indeed, some participants exhibit longer reading times in the repetition condition compared to the inference condition. This variation likely reflects noise inherent in experimental data, not uncommon at all in online reading measures. Such fluctuations are expected in online processing research and do not undermine the main findings, as these cases are distributed across the full range of vocabulary scores rather than clustering at specific points. Thus, while individual errors may introduce some variability, this does not systematically impact the overall relationship between vocabulary and inference cost observed in our study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The authors compare inference to repetition conditions, but do not examine variability due to different inferences or types of inferences. It would be interesting to know if the inferences with the highest reading times had particular natures, syntactic structures, or difficult vocabulary.
Response: This is because there are no different types of inference in the study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
7, 278 The authors state the effect of the complete range of vocabulary scores (35-91) on inference times. These are of course extreme scores; it would be more informative to show the effect over a more modest range, for instance +/- 1 SD.
Response: If we calculate the mean of vocabulary by participants this is equal to 63.31 and the SD is equal to 10.46. Thus, +/- 1 SD around the mean gives a range of 20.92 (min=52.85, max= 73.77). Now, if we calculate the change per 10 unit of vocabulary is equal to 4.46 millisecond, and thus, the change in reading times for a vocabulary range of +/- 1 SD around the mean is equal to 9.34 milliseconds, of 14.01 for +/- 1.5 SD, and 18.68 for +/- 2 SD. We have now added this information to the manuscript.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Figure 3 is impressive, but it gives the false impression (which the authors do not state) that vocabulary scores over 90 would show negligible inference costs. That would only be true for the inferences used here. I realize the authors do not state this, but it might be be helpful to be clear that these results are only for these particular inferences.
Response: Thanks for the suggestion. We have added this information in the discussion section as part of the limitations of the study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Discussion
Some comment on inference difficulty should be made. This could be both in terms of their Inference Cost variable and relative to vocabulary level of the text, and other variables concerning the inferences.
Response: It is difficult to estimate the difficulty of each inference, but we aimed to have relatively easy inferences. All materials are available at https://osf.io/ykfg9/?view_only=0025e1d126c943c49f5f216b482cdcef
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
9, 325 The comment about “deeper” vocabulary knowledge raises questions about the authors’ vocabulary measure. As a receptive measure, it is probably more a measure of vocabulary breadth than vocabulary depth. Investigating the effect of vocabulary depth on inferencing could be suggested as a topic for future research. The authors’ results relate to more “extensive” vocabulary knowledge, not deeper.
Response: While it is true that the vocabulary measure pertains to the breadth of vocabulary dimension, the hypothesis is still open to both dimensions, that is breadth and depth. It is important to notice that a preliminary study we cite used a very similar experimental paradigm in adults, and their vocabulary measure was that of the subtest of vocabulary from the Wechler inventory, and thus, measure vocabulary depth. They also found that better vocabulary was associated with lower inferential processing costs as reflected in reading times. Considering this as part of the understanding we have about the relation between vocabulary and inference making it seems unfair to limit our hypothesis to vocabulary breadth.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Minor corrections
Page 2, line 68 play should be plays
2, 77 “younger” this wording says the authors want research with readers younger than children. Please revise.
3, 136 “did explored” should be “explored”
4, 151 “revised” should be “reviewed”
Response: All corrected. We thank the reviewer for the thorough review of our manuscript.
Reviewer 2
This article sets out to establish whether adolescents with a larger receptive vocabulary process overt and non-overt (inferred) referents at greater speed than adolescents with a smaller receptive vocabulary. After a presentation of a range of previous studies examining the effects of the breadth and depth of vocabulary on comprehension and processing speed in children and adults, this study provides empirical data showing a clear relation between vocabulary knowledge and processing speed in Spanish adolescents.
The introduction, consisting mostly of a (somewhat unordered) literature review, is a little long (146 lines) in comparison with the rest of the article (187 lines), which includes methodology, data and discussion.
The description of the experiment’s setup is rather brief, but links to additional data are provided. The data is generally clearly presented and illustrated in appropriate graphs; the discussion is generally clear and to the point.
However, the analysis of the data has a logical weakness. The author appears to take it for granted that the relation between receptive vocabulary and reading/processing speed is a causal one, but that is not necessarily so; we may well be dealing with a spurious correlation, in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Specific comments and suggested changes
Title: I strongly suggest changing the title to make it clearer, more comprehensible and more attractive to potential readers. The current title is an accurate summary of the findings of the article, but it is quite dense, which makes it difficult to process. Rather than saying that “the vocabulary modulates speed of word integration”, something more general could attract a wider readership, perhaps something along the lines of “Are adolescents with a wider vocabulary quicker at resolving lexical inferences? Evidence from self-paced reading”
Response: What about “Are adolescents with a wider vocabulary faster at inference making during reading? Evidence from self-paced reading”
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Introduction/literature review: As mentioned in the previous section, the introduction consists almost entirely of an overview of related studies. This literature overview lacks order and is, at times, a little repetitive. Thus, in lines 75-82, lines 128-135 and lines 162-165 the author repeatedly observes that there is a lack of studies of adolescents in this area, before going on to summarise yet more existing studies of certain age groups. This section would benefit from being restructured and cleaned up.
Response: We have carefully examined the sections the reviewer pointed out. The first text segment (lines 75-82) offers a general comment on the relation between vocabulary and text comprehension, and the lack of studies in our population of interest. The second segment (lines 128-135) is more specific, referring to the relation between vocabulary and inference making. Only the last segment (lines 162-165) is, in our view, a bit repetitive and thus we have deleted it.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Vocabulary “depth”/ “richness”: There is a brief explanation of what is meant by this (l.38-40), but it is not very clear how this variable can be quantified or whether it is clearly defined and comparable in different studies. As “vocabulary depth” (defined as “the richness of understanding about those words, including semantic associations, morphological characteristics, and syntactic roles”, l.38-40) does not appear to be a variable used in the study presented in this article, that’s not a major problem.
However, throughout the description and discussion of the study, the terms “richer vocabularies” (l.18, 299, 319, 336), “richer vocabulary knowledge” (l. 172, 193, 287) are used, though presumably to refer to the number of words known by participants (i.e. “vocabulary breadth”). This terminological issue should be sorted out.
Response: We thank the reviewer for the detailed scrutiny of the manuscript. Indeed, “richer” was misused in those places the reviewer pointed out. We have corrected all those instances now, changing the phrase “richer vocabularies” for “larger vocabularies” and “richer vocabulary knowledge” for “wider vocabulary knowledge”.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
“Methods” and “Data analysis” sections: Information such as the number of participants is not really part of the method, but simply of the experiment’s setup. However, more information about the participants is needed. In which country was the experiment conducted? (There are native Spanish speakers in many countries.) How were the informants recruited? Did they sign informed consent forms?
Response: The revised manuscript includes this information.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
“Vocabulary” section: This section, consisting of a single sentence, is insufficient. The vocabulary range of the participants in this experiment is the fundamental variable in this study, and the reader should be told how the test works, what kind of lexical items the participants are asked to identify, how many words TEVI-R tests for, etc.; a single reference to Echeverría, Herrera & Segure (2015) is insufficient. More importantly still, the author should explain why this test is the best way of measuring participants’ vocabulary range for this study. It’s a test “standardized for the Chilean population”, but there is no indication in which country the experiment was conducted. TEVI-R can be used for children from the age of 30 months; wouldn’t it be better to use a test tailored specifically for adolescents in order to better determine their vocabulary range?
Response: This section now includes further information about the test and its nature.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Data analysis: This section clearly describes a crucial part of the methodology used in this study, so it should really be part of the Methodology section.
It is not very clear which time intervals the “reading times” refer to. In line 227 it says “we used the reading time TO the first word in the target sentence as control”, but it is not clear what the starting point of the time interval is. Is it the time from the beginning of the context sentence to the first word of the target sentence? In line 228 it says “the reading time FOR the first word”, which doesn’t help to clarify things. In lines 293-295, the author refers to “reading time FOR the word preceding the critical word” (presumably the article), but also “reading times AT the critical word”. These different prepositions are extremely confusing, and it is absolutely essential that the author clarify the exact time intervals that were measured. Is it simply the time it took the respective participant to read the control word/ the critical word?
Response: We apologize for the misuse of the prepositions, which we have corrected in the revised version of the manuscript. Indeed, as a self-paced reading study, any reading time, corresponds to a give word, and therefore we should say “reading times for the target word”, for instance.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
In any case, it seems odd to use the first word in the target sentence as control for the individual’s reading speed, as the first word of the target sentence can be very short, e.g. a definite article, as seen in the example in Table 1, which is unlikely to reveal useful information about the individual’s reading speed because it can be recognized at first sight and doesn’t involve sequential reading as such. Why wasn’t the control reading time measured for a word in the context sentence, or the entire context sentence, where effects of inference or repetition can be ruled out with more confidence?
Response: There were several key reasons for selecting the first word of the target sentence in our experimental materials as the control for reading speed. First, we focused our analysis solely on the data from the target sentence because its structure and content are constant across experimental conditions, which helps to isolate the effects of vocabulary and inference. Second, this initial word showed no significant difference in reading time between conditions, making it a stable baseline for determining overall reading speed. Third, because it immediately precedes the critical word, it serves as the closest temporal baseline just before the experimental manipulation comes into effect. This proximity allows us to control for individual reading speed immediately before the condition effects “kick in.” While it might seem that a short word, such as a definite article, offers limited information about reading speed, its placement provides a reliable measure of transition into the target sentence, capturing individual differences without confounding effects from the varied processing demands in the context sentence. Using a control word from the context sentence, or the entire context sentence, would have introduced additional variability due to the differences in length (between conditions), potentially impacting our baseline reading speed measure.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Table 1/Mixed effects: The data in Table 1 needs a little more explanation, in particular the values given in the last line, “Vocabulary x Condition”. Though the p-value does indicate statistical significance if you take p<0.05 as your threshold, there is little clarity and no discussion about the choice of the threshold(s) for the p-values. This discussion should be added. Can any conclusions be drawn from the differences between the p-values for different variables?
Response: We adopted a p-value threshold of p < 0.05 to determine statistical significance, following conventional practices in cognitive and educational research, and psychology more generally. The significance of the "Vocabulary x Condition" interaction (p = 0.023) suggests a meaningful interaction effect, indicating that the impact of vocabulary on reading times differs across conditions (Repetition vs. Inference). Since interaction between discrete (condition) and continuous (vocabulary) variables are difficult to interpret based only on the output of the regression model we provide two further graphs (Figures 2 and 3). As it can be seen in those plots, the interaction effects implies that vocabulary knowledge more strongly influences reading efficiency when an inference is required, as shown by steeper slopes in the Inference condition. Importantly, this is true for the critical word and not for the baseline.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
References: It is surprising that the author does not mention the following study, which is very similar in scope to the one presented in this article:
Pezoa, José Pablo Pezoa & Pelusa Orellana. 2021. “La relación entre comprensión lectora y vocabulario receptivo en estudiantes chilenos: un estudio exploratorio”, OCNOS Revista De Estudios Sobre Lectura 20 (2), 7-20.
Response: We were aware of the cited work, which is quite different from our study. First, it focused on 5th-graders, just like other studies we cited, not in adolescents. Second, it does not explore the relation between vocabulary and inference making, as we do, but rather on the relation between vocabulary and general reading comprehension. Third, it does not provide any real-time processing measure, such as self-paced reading (as we do), or eye tracking reading. Finally, it is exploratory in nature, and thus, it does not test any particular hypothesis, unlike our study.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Comments on the Quality of English Language
The English language is generally clear and fluent, but with a number of grammatical, lexical and stylistic errors, some of which are listed below. The revised version of this article should be gone over by a native speaker if possible.
l.48, 51: “This means... This suggests...” Style: avoid repetition
l.54: “how vocabulary knowledge influences the cognitive cost” > “how vocabulary knowledge influences cognitive cost”
l.68: “working memory also play a role” > “working memory also plays a role”
l.72: “Other have also found that...” > “Others have also found that...”
l.85: “understand” > “explain”
l.93: “Their findings showed that inferences based on distant information, were significantly more challenging”— Remove comma before “were”.
l.100/101: “this early study... focus on” > “this early study... focused on”
l.136: “In this sense” is a literal translation of the Spanish linker “En este sentido” that cannot be used in the same way in English, as no “sense” has previously been mentioned.
l.136: “did explored” > “did explore” or “explored”
l.150: “underscoring again the pivotal role” > “underscoring, again, the pivotal role” or “again underscoring the pivotal role”
l.200: The mean age (M = 16.2). > The participants’ mean age was just over 16 years (M = 16.2).
l.200-201: There is no need to spell out large numbers. “153 were female, 111 male, and one...”
- 206, 240: “can be found in” > “can be found at”
l.204, 211: “short stories”, “stories” The author should find a different term to refer to these sequences of two sentences; they’re not stories.
l.215: “a critical word positioned always as” > “a critical word always positioned as”
- 216: “we measure” > “we measured”
l.228: “we submitted reading times” Choose a more appropriate verb.
l.231-232: “times below the lower 1% and above the 99%” Rephrase, perhaps “times below 1% and above 99%”
l.233: “we plot” > “we plotted”
l.236: “we compute” > “we computed”
l.239: Full stop after “subject” missing.
l.247: Check the punctuation; one of the periods should probably be a comma.
l.248: “spilled over the three words following the critical” > “spilled over to the three words following the critical one”
l.283: “the study was aimed to investigate” > “the study aimed to investigate”
l.317-318: “vocabulary’s impact” > “the impact of vocabulary”
Response: We profoundly appreciate the work and commitment of the reviewer in proof-reading our manuscript to such depth. We have corrected all errors pointed by the reviewer.
Round 2
Reviewer 1 Report
Comments and Suggestions for Authors
I congratulate the authors on a quick revision that has met all of my previous concerns.
It is good to see that Age was significant as a predictor, and somewhat surprising that it did not reduce the other effects more. This is an important control.
There were several small points:
line 240 should be participant's
line 289 (Table 1) This should be Table 2
The p and significance columns are redundant - I suggest using one or the other. In the p column, one should not say that p = .000 (i.e., 0).
line 294 should be range
Author Response
Dear Reviewer 1, below you will find a detailed response to all your comments, and those to Reviewer 2. Again, I opted for provinding responses to both of you. In the revised manuscript, you will find all changes marked in blue font.
Response to Reviewers: Second Round
Reviewer 1
I congratulate the authors on a quick revision that has met all of my previous concerns.
It is good to see that Age was significant as a predictor, and somewhat surprising that it did not reduce the other effects more. This is an important control.
There were several small points:
line 240 should be participant's
line 289 (Table 1) This should be Table 2
The p and significance columns are redundant - I suggest using one or the other. In the p column, one should not say that p = .000 (i.e., 0).
line 294 should be range
Response: We thank the reviewer for their positive evaluation of our manuscript. All minor changes made in this second round have been addressed.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Reviewer 2
The authors have still not addressed the main logical weakness of their analysis, namely that the relation between receptive vocabulary and reading/processing speed in not necessarily a causal one; we may well be dealing with a spurious correlation in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
Response: We apologize for not having addressed the comment in the previous round of reviews. In the current version we have addressed the issue that the reviewer is pointing to in the discussion, explicitly saying that this relation might not be causal.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The authors' response to my question as to why the first word of the target sentence, often a short word like a definite article, was used as control for the individual’s reading speed is well argued, but this explanation should appear in the article, not just in the authors' response to the referee report. It's an obvious question that many readers will ask, so it would be better to provide this explanation to all readers.
Response: We have now included our rationale for using that region as control explicitly in the text.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The same goes for the explanation/justification of the p-value threshold, which should appear in the article. Explanations given in the response to reviewers don't make the article clearer to its readers once it's published.
At least these three issues should be addressed in the article before it is published.
Response: We have now included our rationale for using a threshold of p values at .05 explicitly in the text.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Comments on the Quality of English Language
Some of the changes made to the text require corrections, among others:
- 200: "in social media platforms" > "on social media platforms"
- 234-254: The inconsistent use of present and past tense should be sorted out: looked at (past), we look at (present), we use (present), we included (past)...
- 240: "participants age" > "participants' age" or "participant age"
- 246: "within-subject designs": What does this mean? Rephrase.
- 257/258: "as it can be seen in the figure" > "as can be seen in the figure"
l.2982: "there were main effects": rephrase; "main" doesn't fit here.
- 283/284 "were faster readers: of vocabulary" > "were faster readers of vocabulary"
l.293/294: "the effect over a more modest vocabulary ranges, i.e.. +/- SD..." Check grammar and expression: "vocabulary range"? “the effect for participants with a more modest vocabulary range”?
Response: Again, we deeply appreciate the work and commitment of the reviewer in proof-reading our manuscript to such depth. We have corrected all errors pointed by the reviewer or modify parts of the text around those errors to make the manuscript clearer. All our changes are marked in blue font.
Reviewer 2 Report
Comments and Suggestions for Authors
The authors have still not addressed the main logical weakness of their analysis, namely that the relation between receptive vocabulary and reading/processing speed in not necessarily a causal one; we may well be dealing with a spurious correlation in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
The authors' response to my question as to why the first word of the target sentence, often a short word like a definite article, was used as control for the individual’s reading speed is well argued, but this explanation should appear in the article, not just in the authors' response to the referee report. It's an obvious question that many readers will ask, so it would be better to provide this explanation to all readers.
The same goes for the explanation/justification of the p-value threshold, which should appear in the article. Explanations given in the response to reviewers don't make the article clearer to its readers once it's published.
At least these three issues should be addressed in the article before it is published.
Comments on the Quality of English Language
Some of the changes made to the text require corrections, among others:
l. 200: "in social media platforms" > "on social media platforms"
l. 234-254: The inconsistent use of present and past tense should be sorted out: looked at (past), we look at (present), we use (present), we included (past)...
l. 240: "participants age" > "participants' age" or "participant age"
l. 246: "within-subject designs": What does this mean? Rephrase.
l. 257/258: "as it can be seen in the figure" > "as can be seen in the figure"
l.2982: "there were main effects": rephrase; "main" doesn't fit here.
l. 283/284 "were faster readers: of vocabulary" > "were faster readers of vocabulary"
l.293/294: "the effect over a more modest vocabulary ranges, i.e.. +/- SD..." Check grammar and expression: "vocabulary range"? “the effect for participants with a more modest vocabulary range”?
Author Response
Dear Reviewer 2, below you will find a detailed response to all your comments. In the revised manuscript, you will find all changes marked now in blue font.
Response to Reviewers: Second Round
Reviewer 2
The authors have still not addressed the main logical weakness of their analysis, namely that the relation between receptive vocabulary and reading/processing speed in not necessarily a causal one; we may well be dealing with a spurious correlation in which both vocabulary knowledge and reading/processing speed are dependent on a third factor, such as reading practice. Adolescents who read more are likely to have a more developed vocabulary, and they are also likely to have developed more efficient cognitive processes that allow them to read and infer more quickly, without the former being the cause of the latter. This possibility should, at the very least, be explored in the discussion.
Response: We apologize for not having addressed the comment in the previous round of reviews. In the current version we have addressed the issue that the reviewer is pointing to in the discussion, explicitly saying that this relation might not be causal.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The authors' response to my question as to why the first word of the target sentence, often a short word like a definite article, was used as control for the individual’s reading speed is well argued, but this explanation should appear in the article, not just in the authors' response to the referee report. It's an obvious question that many readers will ask, so it would be better to provide this explanation to all readers.
Response: We have now included our rationale for using that region as control explicitly in the text.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
The same goes for the explanation/justification of the p-value threshold, which should appear in the article. Explanations given in the response to reviewers don't make the article clearer to its readers once it's published.
At least these three issues should be addressed in the article before it is published.
Response: We have now included our rationale for using a threshold of p values at .05 explicitly in the text.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Comments on the Quality of English Language
Some of the changes made to the text require corrections, among others:
- 200: "in social media platforms" > "on social media platforms"
- 234-254: The inconsistent use of present and past tense should be sorted out: looked at (past), we look at (present), we use (present), we included (past)...
- 240: "participants age" > "participants' age" or "participant age"
- 246: "within-subject designs": What does this mean? Rephrase.
- 257/258: "as it can be seen in the figure" > "as can be seen in the figure"
l.2982: "there were main effects": rephrase; "main" doesn't fit here.
- 283/284 "were faster readers: of vocabulary" > "were faster readers of vocabulary"
l.293/294: "the effect over a more modest vocabulary ranges, i.e.. +/- SD..." Check grammar and expression: "vocabulary range"? “the effect for participants with a more modest vocabulary range”?
Response: Again, we deeply appreciate the work and commitment of the reviewer in proof-reading our manuscript to such depth. We have corrected all errors pointed by the reviewer or modify parts of the text around those errors to make the manuscript clearer. All our changes are marked in blue font.
Round 3
Reviewer 2 Report
Comments and Suggestions for Authors
The authors have now added the the most important information that I felt was necessary for this article to be published. Thanks for taking my advice into consideration.
