WordDriver

WordDriver is based on the evidence supporting the dual-route model of skilled reading (Coltheart, 2006), Ehri's phase model of reading development (Ehri, 2005), and Share's phonological recoding theory (Share, 1995). Research examining these theories has shown that:

• Accurate word reading is an essential component of skilled reading - the simple view of reading (Gough & Tunmer, 1986; Lervag, Hulme, & Melby-Lervag, 2018).
• Skilled readers use two ways of reading words: automatic recognition of known words, and sounding out and blending (decoding) for unfamiliar words - the dual route model of skilled reading (Coltheart, 2006).
• Accurate decoding has been shown to result in development of clear mental images of words (orthographic representations) which leads to automatic word recognition and reading fluency - the Phonological Recoding theory (Share, 1995). Further, formation of orthographic representations is enhanced when children decode words which are presented in isolation, and are provided with corrective feedback (Martin-Chang, Ouellette, & Bond, 2017).
• Some children have specific difficulties acquiring the skills that underpin accurate word reading. They may have problems mastering the decoding process (often due to weak phonological processing skills), and/or they may take longer to develop orthographic representations (Snowling & Hulme, 2012).
• Other research has shown that these children eventually master accurate decoding, but they take longer and require more repetitions to develop orthographic representations (Apel, Thomas-Tate, Wilson-Fowler, & Brimo, 2012).
• WordDriver uses many of these evidence-based features: it randomly presents items (words and nonwords) in isolation and encourages extensive decoding practice at different levels of difficulty. It is delivered in an individual situation allowing the instructor to provide corrective feedback to optimise formation of orthographic representations of words, and encourage links between the word, its pronunciation, and its meaning.

Our preliminary research suggests that this intervention has the potential to be an efficient evidence-based component of reading interventions for children with severe and persistent word reading delay (Seiler et al, 2019 ). Using single subject research designs, we found that all participants (aged 7-8 years who had not responded to previous reading interventions) made significant gains in decoding skills within 15 x 20 minutes sessions.

WordDriver displays graphics on the screen that use an analogy of learning to drive a car. A functional version is being made available without charge for clinicians and other researchers to examine and perhaps integrate into their own projects. We would appreciate feedback and discussion about any aspects of WordDriver and this project - please email info@elr.com.au

WordDriver has two stages:

1. WordDriver-1 provides training in accurate decoding. It is designed for children who have mastered grapheme-phoneme (letter-sound) knowledge of short vowels and single consonants, but are unable to use this knowledge to accurately decode short words (3-letter, 4-letter words). It presents items with 1:1 letter-sound correspondence (starting at 2- and progressing to 6-letter items).
2. WordDriver-2 follows on from WordDriver-1. It aims to expand orthographic knowledge by delivering items with consonant and vowel digraphs.

WordDriver modules

• An administrative "Loader module" allows the instructor to select appropriate intervention and testing modules to match each child's decoding intervention needs.
• Intervention modules: In both WordDriver-1 and WordDriver-2, the learner progresses from an L-Plate (learning), to a P-Plate (practising), and then a D-Plate (driver), each presenting a wide range of target items: words and nonwords drawn from the MRC and ARC databases. For instance, in the 4-letter word D-Plate in WordDriver-1, decoding targets are drawn from a pool of 468 items (234 words, each with a nonword matched for orthographic similarity, eg flat - clat).
• Testing modules: Both stages offer T-Plates (testing) which enable the instructor to gather pre-intervention decoding accuracy and monitor progress over time. WordDriver-1 also includes an S-Plate (speed). This module was used in our initial research to measure changes in motor response to control for any increased rate when using the app. We found no significant change over time, hence improved skill at using the technology was ruled out as a factor in improved rate of decoding.

Within the intervention modules (L-Plate, P-Plate, D-Plate):

• The Go button displays an item (a randomly presented word or nonword). The items increase in difficulty based on orthographic complexity (starting with frequent and progressing to less frequently occurring letter combinations).
• The student (or the instructor in the case of the L-Plate) decodes the item and is provided with corrective feedback.
• The instructor clicks the Correct button following an accurate response, and the Help button for an inaccurate response. There are three levels of Help: visual highlighting of the letters/graphemes to prompt accurate decoding, visual plus auditory feedback to demonstrate decoding, and, finally, the instructor touches each letter/grapheme and verbally performs decoding. Note: it is recommended that only the first level of Help be used for accents other than Australian.
• To support development of automatic word recognition, the instructor/student uses words in a sentence to illustrate the meaning.
• The student clicks/touches the Book icon for words or the Bin icon for nonwords.
• The student clicks/touches the Go button for the next item.

Within the testing modules (T-Plate):

• The instructor prints a response form for the selected T-Plate. This enables the instructor to record accuracy of decoding for each item.
• The student touches the Go button to display a nonword. All nonwords used within WordDriver are "legal", that is, the grapheme combinations are consistent with the English language, and no word or nonword is repeated between any of the modules.
• The student reads the item (the nonword).
• The instructor records accuracy of response on the printed response form. No corrective feedback is provided.
• The student touches the Go button to reveal the next and subsequent items.
• The instructor tells the student to stop after six errors within eight items has occurred.

Storage of results:

• At any time after starting (and before terminating) any of the intervention or testing modules, the instructor may click either the number plate (bottom right of the screen), or the licence plate (bottom left), to see a detailed analysis of the results.
• A log of completed results is available from the "Tools" section of the administrative Loader module, and may be copied or deleted at any time.

From the loader module

• Choose the number plate, eg AMS001
• The intervention (L-Plate, P-Plate or D-Plate) or testing module (T-Plate)
• And the level of difficulty, eg 2-letter, 3-letter, 4-letter

Intervention: L-Plate, P-Plate, D-Plate

All modules are delivered as described in the above overview, except the L-Plate. The L-Plate occurs in WordDriver-1 to familiarise the student with the software, and to teach the decoding process at each level (i.e., 3-letter, 4-letter, and so on). As a teaching module, it differs from the P-Plate and D-Plate in two ways. First, the instructor does all actions: touching the Go button, demonstrating decoding, and putting the items in the Book or Bin. Second, the Help button is used for all items as this allows increased opportunity for the instructor to teach the decoding process - sounding out and blending.

Following pre-intervention assessment, the instructor determines an appropriate starting level (2-letter, 3-letter etc.). Progression to the next level (eg from 3- to 4-letter items) occurs once the student achieves 90% accuracy at the current level. For example, if a student is on 3-letter words, a score of 90% on the P-Plate allows progression to the D-Plate; and a score of 90% suggests the student can move to the 4-letter level.

Testing: T-Plate

There are 41 nonword lists, each comprising items with 1:1 letter-sound correspondence. To deliver the T-Plate, the instructor selects the list and the length (70 or 35 items), and the T-Plate is administered as described above. Complete nonword lists/response forms for these T-Plates are available here.

From the loader module

• Choose the number plate, eg ABC001
• Select the focus: consonant digraphs or vowel digraphs
• And then the licence plate (T-Plate, P-Plate, D-Plate). There are no L-Plates in WordDriver-2

Intervention: P-Plate and D-Plate

On selection of a P-Plate and D-Plate, you are presented with a choice of targets. This allows you to provide specific intervention for one or a group of digraphs.

Following intervention of the selected target digraphs, you may wish to then include "Foil" words (words and nonwords which contain only short vowels), as this develops discrimination skills. For example, to accurately decode "rat" versus "rate", or "hut" and "hurt".

Testing: T-Plate

Each T-Plate currently allows assessment of a group of digraphs. This allows the instructor to measure change in the targeted compared to untreated digraphs. Delivery is as described above. Complete nonword lists/response forms for these T-Plates are available here.

WordDriver can be shared in teletherapy sessions as a "shared screen" in Zoom, and as the "Add-On" WordDriver for Coviu. On either platform, both presenter and the client can interactively share the mouse/touch controls.

Two, free public access modes are available

Anyone may use the "Unregistered" mode which has only 2 predefined "number plates" ("ABC123" and "XYZ789"). The "Registered" mode is more suitable if you wish to use WordDriver for your research or teaching, and requires that you have arranged with the authors for your own (free) set of "number plates".

• WordDriver-1
1. Unregistered access
2. Registered users only
3. Nonword lists/response forms for T-Plate
• WordDriver-2
1. Unregistered access
2. Registered users only
3. Nonword lists/response forms for T-Plate

Usage and privacy notes

• WordDriver will run in any "modern" browser (for example Chrome, Microsoft Edge or Safari) on computers and tablets. On computers, it's usually advisable to operate in "full screen" mode (eg press F11).
• No usage data is saved or transmitted outside your browser. A detailed analysis of results in a module is maintained only during its operation. A log of completed results is available from the "Tools" section of the administrative Loader module, and may be deleted at any time.

1. PDF Preliminary Study (2013) Journal of Clinical Practice in Speech-Language Pathology
2. Link to Download Thesis (2015) by Dr. Toni Seiler from the Curtin University institutional repository
3. PDF Summary of Thesis (2017) by Toni Seiler describing the WordDriver research project
4. PDF Unpublished Paper (2017) describing WordDriver-1 design and development
5. Link to Research Report (2019) International Journal of Language and Communication Disorders
6. PDF Literacy and WordDriver 1 (2020) Learning Difficulties Australia: Developing accuracy and fluency in word reading skills
7. PDF Literacy and WordDriver 2 (2023) Australian Journal of Learning Difficulties: Targeting phonological recoding to support orthographic learning: effectiveness of WordDriver delivered via telehealth
8. PDF Nonword Assessment Response Forms (2020-23) As used in WordDriver 1 & 2
9. Video Short clips showing WordDriver in use

1. Apel, K., Thomas-Tate, S., Wilson-Fowler, E. B., & Brimo, D. (2012). Acquisition of initial mental graphemic representations by children at risk for literacy development. Applied Psycholinguistics, 33(2), 365-391. doi:10.1017/s0142716411000403
2. Coltheart, M. (2006). Dual route and connectionist models of reading: An overview. London Review of Education, 4(1), 5-17. doi:10.1080/13603110600574322
3. Gough, P. B., & Tunmer, W. E. (1986). Decoding, reading, and reading disability. Remedial and Special Education, 7, 6 - 10.
4. Lervag, A., Hulme, C., & Melby-Lervag, M. (2018). Unpicking the developmental relationship between oral language skills and reading comprehension: It's simple, but complex. Child Development, 89(5), 1821-1838. doi:doi:10.1111/cdev.12861
5. Martin-Chang, S., Ouellette, G., & Bond, L. (2017). Differential effects of context and feedback on orthographic learning: How good Is good enough? Scientific Studies of Reading, 21(1), 17-30. doi:10.1080/10888438.2016.1263993
6. Seiler, A., Leitao, S., & Blosfelds, M. (2018). WordDriver-1: Evaluating the efficacy of an app-supported decoding intervention for children with reading impairment. International Journal of Language & Communication Disorders, http://dx.doi.org/10.1111/1460-6984.12388(0). doi:doi:10.1111/1460-6984.12388
7. Seiler, A. & Leitao, S. (2023) Targeting phonological recoding to support orthographic learning: effectiveness of WordDriver delivered via telehealth, Australian Journal of Learning Difficulties, 28:1, 1-26, DOI: 10.1080/19404158.2023.2208135
8. Share, D. L. (1995). Phonological recoding and self-teaching: sine qua non of reading acquisition. Cognition, 55(2), 151-218. doi:10.1016/0010-0277(94)00645-2
9. Snowling, M. J., & Hulme, C. (2012). Annual Research Review: The nature and classification of reading disorders - a commentary on proposals for DSM-5. Journal of Child Psychology and Psychiatry, 53(5), 593-607. doi:10.1111/j.1469-7610.2011.02495.x
10. Rinderknecht, M.D., Ranzani, R., Popp, W.L. et al. Algorithm for improving psychophysical threshold estimates by detecting sustained inattention in experiments using PEST. Atten Percept Psychophys 80, 1629-1645 (2018). https://doi.org/10.3758/s13414-018-1521-z