In short
- Write down the problem first, then put the same twelve questions to every vendor.
- The texts come first: a platform whose stories your pupils cannot read, or cannot read enough of, fails whatever else it offers.
- Teacher time, data protection and cost per pupil per year are checks in their own right, not details for later.
- Ask to see each answer working on screen, score it on the spot, and finish with a trial in one class.
Ed-tech is often bought on thin evidence. Morrison, Ross and Cheung (2019) studied how ed-tech products were bought, with people from 54 school districts and 47 companies. Needs assessments were rarely conducted. Decisions often rested on small pilots and peer references, and less often on rigorous evidence. A checklist cannot fix that alone, but it makes you ask the same questions of every product. Below are twelve checks, why each matters, the question to put to the vendor, and how to score the answers.
How do you choose a reading platform for your school?
Choose a reading platform in five steps. Write down the problem it must solve. Put the same twelve checks to every vendor, in writing and then on screen. Shortlist two or three products. Then try the best one with one class before you buy.
- Name the problem. One sentence, with the classes and a number from your own records.
- Send the checklist. The same twelve questions to every vendor, so the answers can be compared.
- Watch it on screen. Ask the vendor to show each answer working, not describe it.
- Shortlist two or three. Drop any product that fails a check you marked as essential.
- Try one with a class. Four to six weeks, with a pass mark written in advance.
The first step matters most. The Education Endowment Foundation's guide to implementation (Sharples, Albers, Fraser and Kime, 2019) advises schools to define the problem they want to solve before they look at programs. It warns that considering new approaches before examining the problem will bias your view of it.
'Our pupils should read more' is a wish. 'In Grade 6, fewer than half the pupils finished a book last term' is a problem you can test a platform against. That example is invented; use your own figures.
The same guide asks of any program: how strong is the evidence, how easy will it be to implement, and is it cost effective? The twelve checks make those questions concrete.
The twelve checks at a glance
Use the table as your list. The last column is the question to send the vendor.
| Check | Why it matters | Question to ask the vendor |
|---|---|---|
| 1. Levelled texts | Pupils cannot read for meaning if every line has unknown words | Which level system do you use? Can we see a sample at each of our levels? |
| 2. Amount of text | A term of weekly reading needs many texts at each level | How many texts are there at each of our levels, and how long is each? |
| 3. Audio | Some pupils need to hear the text while they follow it | Which texts have audio, is it synced to the text, and can pupils turn it off? |
| 4. Checks of understanding | Short questions show who followed the text | What do pupils answer after a text, and how long does it take? |
| 5. Word practice | Words from the texts need to come back if pupils are to keep them | How do the words from each text come back later, and who sets that up? |
| 6. Teacher records | You cannot help a pupil you cannot see | Can a teacher see, this week, who has read what and who has stopped? |
| 7. Leader reports | Coordinators and boards need class and grade totals | What can a coordinator see by class and grade, and can it be exported? |
| 8. Teacher time | Work that lands on teachers is a cost the invoice does not show | What must a teacher set up, mark or enter by hand each week? |
| 9. Data protection | Pupils' data needs a written promise from the vendor | What do you collect, who sees it, where is it kept, and what happens when we leave? |
| 10. Cost per pupil per year | Quotes cannot be compared until they are in the same unit | What is the full cost per pupil per year, with renewal, add-ons and training? |
| 11. Support and training | Teachers need help after the first week, not only at the start | What training is included, and who helps a teacher in week five? |
| 12. A trial | Real pupils tell you more than a demonstration | Can we try it with one class for four to six weeks before deciding? |
Score each check as you go: present, partial or missing. Before the first call, mark three or four checks as essential. A product that misses one is out, however well it does on the rest. This is practice, not a research finding.
Checks 1 to 5: the reading itself
The texts are the product. Good reports cannot rescue stories your pupils cannot read.
Levelled texts
Ask for a sample text at each level you teach, and test it on pupils, not on yourself. Give three or four pupils a page and ask them to mark the words they do not know. To place pupils in the first place, see How to Assess a Student's English Reading Level.
Level matters for software too. Escueta, Nickow, Oreopoulos and Quan (2020) reviewed experimental studies of education technology in developed countries. Writing about learning software, mostly in mathematics, they suggest that matching pupils' actual learning levels may be an important factor in whether it improves learning.
Amount of text
Count the texts at each level against the weeks in your term. A library with a large total can still hold only a handful of texts at the level of your weakest class. What Should a Reading Pack Contain? works out how many texts a term needs.
Audio, checks of understanding and word practice
Audio lets a pupil follow the words while a voice reads them. Ask which texts have it, and whether it is optional.
For checks of understanding, time one yourself during the demonstration. A quiz that takes longer than the reading has the balance the wrong way round.
For word practice, ask where the words come from. Practice built from the texts the class has just read brings those words back.
Checks 6 to 8: what teachers see, and what it costs them
A record is useful only if a teacher opens it. Ask the vendor to log in as a teacher and show you which pupils have not read this week. Then ask for the coordinator's view by class and grade, and whether it exports to a file you can share.
Then ask about teacher time. What must a teacher set up each week? What must be marked by hand? What happens to a class when its teacher is absent?
Kraft (2020) notes that effects tend to shrink when small programs are taken to scale. He warns that a program is unlikely to keep its effect at scale if it needs more skill than typical teachers have. Depending on one charismatic person is another warning sign. So ask each question with your busiest teacher in mind, not your keenest one.
Check 9: data protection
Ask the data questions before any pupil gets an account, including the pupils in a trial. If your school is covered by the European Union's General Data Protection Regulation (GDPR, Regulation (EU) 2016/679), two articles frame the check.
- Article 5 says personal data must be adequate, relevant and limited to what is necessary for the purposes of processing. Ask for the list of data collected on each pupil, with the reason for each item.
- Article 28 covers a company that processes data on behalf of another body, the controller. The controller must use only processors that give sufficient guarantees of appropriate technical and organisational measures. The processing must be governed by a contract or other legal act.
Whether the GDPR applies to your school, and who is the controller, are questions for your adviser. Student Data Privacy: Questions to Ask an EdTech Vendor gives seven questions to send, with good and worrying answers for each.
Checks 10 to 12: cost, support and a trial
Cost per pupil per year
Ask every vendor for the same figure: the full cost per pupil per year. It should include the licence, any add-ons, training and the price at renewal. Then add your own staff hours, which no quote includes. What Should a School Reading Program Cost? A Budgeting Guide shows how to build that figure.
Support and training
Ask what happens after the first training session. Sharples et al. (2019) describe a common mistake: providing only up-front training, with little or no follow-on support. Ask who a teacher contacts in week five.
A trial
Ask for a trial with one class before any purchase. It shows what a demonstration cannot: whether pupils use the product when nobody from the vendor is watching. How to Run a Pilot of a Reading Platform explains how to set a pass mark and decide on the result.
What to make of a vendor's evidence
Read a vendor's results with care. Sharples et al. (2019) advise schools to treat a claim of huge impact as a warning sign.
Realistic effects are modest. Cheung and Slavin (2013) reviewed 20 studies of educational technology for struggling readers in grades 1 to 6, with about 7,000 pupils. The effect on reading skills was positive but small (effect size 0.14). Those pupils were struggling readers in primary school, so use the figure only as a sense of scale.
Size matters too. Kraft (2020) found that studies with 100 pupils or fewer report far larger effects than studies with more than 2,000. A striking result from one school is weak evidence on its own.
Four questions sort strong evidence from weak:
- How many pupils took part, and of what age and level?
- What were they compared with: a similar class, or the same pupils before they started?
- Who ran the study, and was it published?
- Was the outcome a broad reading test, or a test of the product's own content?
If a vendor has no independent study, note it, and let your own trial carry more weight.
How to run the checklist with a vendor
Send the twelve questions by email before the demonstration, so the answers are in writing. On the call, ask the vendor to show each answer on screen. A written answer can describe a plan. A screen either shows the thing working or it does not. This is practice, not a research finding.
Example (invented school)
An English coordinator marks four checks as essential: levelled texts, amount of text, teacher records and a trial. Vendor A has only two texts at A1, where her Grade 4 classes read, so A is out. B and C pass every essential check. B scores partial on teacher time, because its quizzes are marked by hand. She asks the head for a six-week trial of C, with B as the fallback.
Keep the score sheets for the meeting after the trial. They record what each vendor promised.
How iRead handles this
In iRead, every story, quiz and game is levelled to the CEFR, from A1 to C2, and many stories have an audio read-along synced to the text. A quiz runs after every session, and four word games are regenerated every 24 hours from the story's vocabulary, with no teacher setup. Teachers see retention per word, per game and per class. Individual student data is visible only to the school's own staff. A school can run a free pilot with one class for 4 to 6 weeks before any school-wide decision.
The core principle: Ask every vendor the same twelve questions and ask to see each answer on screen. Then let a trial with one class decide.
See it in iRead: Every iRead story, quiz and game is levelled to the CEFR, from A1 to C2. A school can run a free pilot with one class for 4 to 6 weeks before any school-wide decision.
See the iRead programKey takeaway
Write the problem first, ask every vendor the same twelve questions, and score what you see. A product that fails on the texts fails, whatever else it offers.
Frequently asked questions
What features should a reading platform have?
Texts at the levels you teach and enough of them for a term. Audio for pupils who need it, short checks of understanding, practice on the words from the texts, and a record teachers can read. Around those, check data protection, cost per pupil per year, support and the chance of a trial.
How do I choose a reading platform for my school?
Write down the problem first, with a number from your own records. Send the same questions to every vendor, ask to see each answer on screen, and shortlist two or three. Then try the strongest with one class for four to six weeks before you decide.
Which checks matter most?
The texts. If your pupils cannot read the stories, or there are too few at their level, nothing else can rescue the product. After that, weigh teacher time and teacher records, because they affect whether a busy teacher keeps using it. Mark your own essential checks before the first call.
When should we ask a vendor about student data?
Before any pupil gets an account, including the pupils in a trial. Ask what is collected on each pupil and why, who can see it, where it is stored and for how long, and what happens to it when the contract ends. Get the answers in writing.
Sources
- Sharples, J., Albers, B., Fraser, S., & Kime, S. (2019). Putting Evidence to Work: A School's Guide to Implementation. Guidance Report (2nd ed.). Education Endowment Foundation, London. d2tic4wvo1iusb.cloudfront.net/eef-guidance-reports/implementation/EEF_Implementation_Guidance_Report_2019.pdf
- Morrison, J. R., Ross, S. M., & Cheung, A. C. K. (2019). From the market to the classroom: How ed-tech products are procured by school districts interacting with vendors. Educational Technology Research and Development, 67(2), 389-421. doi.org/10.1007/s11423-019-09649-4
- Escueta, M., Nickow, A. J., Oreopoulos, P., & Quan, V. (2020). Upgrading education with technology: Insights from experimental research. Journal of Economic Literature, 58(4), 897-996. doi.org/10.1257/jel.20191507
- Kraft, M. A. (2020). Interpreting effect sizes of education interventions. Educational Researcher, 49(4), 241-253. doi.org/10.3102/0013189X20912798
- Cheung, A. C. K., & Slavin, R. E. (2013). Effects of educational technology applications on reading outcomes for struggling readers: A best-evidence synthesis. Reading Research Quarterly, 48(3), 277-299. doi.org/10.1002/rrq.50
- European Parliament and Council of the European Union (2016). Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation). Official Journal of the European Union, L 119, 4.5.2016, 1-88. eur-lex.europa.eu/eli/reg/2016/679/oj