Release notes

Summary of changes to the software by version

Note on version numbers

The 3di installer has a filename like 3di5_uk_230122_setup.exe. This filename includes the version number 230122 of the software in yymmdd format, that is year month day, so in our example 22 January 2023. Sometimes you will see a suffix x which denotes a sub revision.

Version numbers of this form are also used in the documents provided in the PreEntry and Reference subfolders of your 3di installation. Sometimes the version numbers of these documents are much older than your 3di software — it doesn't mean the document is out of date, just that it hasn't needed any changes in a while.

Changes

This section lists all changes introduced to the 3di, organised by version in reverse chronological order (newest first).

If you are upgrading from an earlier version, read the entries below starting with the first version after the one you have been using, and continue upwards.

Changes introduced in 251219a

Summary

This release fixes a single minor scoring issue.

#133 DSM-5 extended question 713 reversed scoring

Fixes inverted scoring of question 713 (8.10.2) in the DSM-5 ASD extended assessment.

This question reads:

And now I'd like us to look briefly at Sam's relationships with others. Would you say that: - she has at least one or two good friends

• A response of "definitely applies" was incorrectly treated as the more symptomatic option, inflating the A3d Developing and maintaining relationships - Making friends score by 9%.

• A response of "no, does not apply" was incorrectly treated as the less symptomatic option, producing no increase in the subscale score when it should raise it by 9%.

This fix reverses the scoring of those two options to produce corrected subscale percentages.

Note that in the DSM-5 ASD extended assessment, we recommend you always complete every question on the route before generating the report. If you do not, the above errors could be amplified, since questions you did not answer are scored by averaging the answers from those you did.

Changes introduced in 251219

Summary

Changes have been made which correct and improve the scoring of the DSM-5 ASD brief assessment. This is accompanied by revised guidance on interpretation of report scores, communicated to users separately. This includes a tightening of eligibility criteria and the requirement that every question in the assessment must be assigned a scorable response in order to compute percentage scores.

Minor corrections have also been made to the scoring of the DSM-5 ASD extended assessment.

Small changes have been made to clarify the user interface and address usability bugs.

Clinically relevant changes

These changes relate to the clinical content of the 3di software and alter the interview content, scoring or reporting in some way.

#78 DSM-5 brief question 317 reversed scoring

Fixes an issue in the DSM-5 ASD brief assessment where inverted scoring was applied for question 317 (12.2.4) "Does Sam vary her behaviour appropriately according to whom she is with or where she is?". The "never" response was scored as asymptomatic with 0 points, while the "often" response was scored as symptomatic with 2 points. The scoring is now corrected.

For a full description, see alg4 errata November 2025.

#79 DSM-5 brief "with social impact" questions are scored too low

Fixes an issue in the DSM-5 ASD brief assessment where the "with social impact" response for six questions in Criterion B incorrectly led to no increase in the subscale percentage scores.

For a full description, see alg4 errata November 2025.

#81 Notes may be missing from report

Fixes an issue in the feature which allows you to record written notes on individual questions while conducting the interview. Not all notes relevant to an assessment appeared underneath the relevant scoring tables in the report. Some notes would appear twice. Now, any notes taken on a question in one of the four ASD assessments appear exactly once underneath the scoring table for that assessment.

#93 Assign distinct score to "with social impact" responses

Fixes an issue where several questions that had a "with social impact" response were not scored optimally. This change is in addition to #79 above.

In the DSM-5 ASD brief assessment, each scorable question contributes equally to its subscale. Responses are mapped linearly to values between 0 and 1, averaged, and expressed as a percentage.

Seventeen questions under Criterion B offer four response levels, the highest being "with social impact". Previously, the top two response levels were both mapped to the maximum score, while the second level ("minimal or transient") was mapped to half of that, and the lowest level to zero. Therefore although four response options were available to the clinician, only three distinct scoring outcomes existed.

To restore consistency with the rest of the instrument, all four levels now receive distinct scores, evenly spaced across the 0-1 range (0, ⅓, ⅔, 1). According to our research as communicated in alg4 errata November 2025, this produces a small but measurable improvement to the sensitivity and specificity of the instrument to predict autism diagnoses.

#94 DSM-5 brief should report applicability warnings

The applicability criteria for the DSM-5 ASD brief assessment have been formalised:

1. The child must be at least 5 years old at time of interview.

2. The child must have at least fluent speech, as recorded in question 8.1.1.

3. All of the 63 scorable questions of the route ASD_brief (DSM-5 Crit A+B) must have been completed with scorable responses in order to generate the scoring table in the report.

A warning now appears underneath the scoring tables if either of condition 1 or 2 is not met.

If condition 3 is not met, the scoring tables do not appear at all. Since the DSM-5 ASD brief assessment is already of reduced length, we consider it essential to answer every question with a scorable response.

#95 DSM-5 extended question 729 reversed scoring

Fixes inverted scoring of question 729 (8.9.11) in the DSM-5 ASD extended assessment.

This question reads:

I'd like to ask a few questions about the way in which Sam relates to others in conversation. Would you say that: - she makes good use of gestures to get her meaning across?

• A response of "definitely applies" was incorrectly treated as the more symptomatic option, inflating the A2e Non-verbal social communication - Body language score by 7%.

• A response of "no, does not apply" was incorrectly treated as the less symptomatic option, producing no increase in the subscale score when it should raise it by 7%.

This fix reverses the scoring of those two options to produce corrected subscale percentages.

Note that in the DSM-5 ASD extended assessment, we recommend you always complete every question on the route before generating the report. If you do not, the above errors could be amplified, since questions you did not answer are scored by averaging the answers from those you did.

#96 DSM-5 extended B1a scores are confined to range 0-73%

Fixes the scoring of the "unable to judge" response to 12 questions in B1 Stereotyped behaviours - Stereotyped and repetitive speech in the DSM-5 ASD extended assessment.

In the DSM-5 ASD extended assessment, every scorable question takes responses from among 2, 3 or 4 levels of symptom severity or frequency. The subscale percentage score is computed by mapping each response to a number between 0 to 1, taking the average of these, then expressing the result as a percentage.

The "unable to judge" option is intended to be unscorable: if selected, the question should be excluded and its contribution replaced by the mean of the remaining scorable items.

Previously, "unable to judge" was incorrectly treated as a fourth, maximum level of severity. "definitely applies" should have been the maximum level, but a child with every "definitely applies" response selected would achieve a score of only 73% in Stereotyped and repetitive speech.

The fix restores the correct behaviour:

• "unable to judge" is treated as an unscorable response and excluded from the calculation.

• The remaining three levels are scored on the usual 0–½–1 mapping.

A maximally symptomatic child with all scorable responses will now correctly receive 100% for this subsubscale.

#99 Update wording of q99

Alters the wording of question 11.8.9 in interview subsection 11.8: Negative sensory reactions. This reflects that hand dryers are nowadays a more common source of loud noise than vacuum cleaners.

#100 Reword four CCC questions

Four questions in interview section 8: Language and non-verbal communication have minor changes to their wording to bring them up to date. Specifically, the wording of the following has been updated: question 698 (8.8.2), question 701 (8.8.5), question 711 (8.9.14), question 723 (8.11.3).

#102 DSM-5 extended question 1227 reversed scoring

Fixes inverted scoring of question 1227 (10.3.7) in the DSM-5 ASD extended assessment.

This question reads:

Does Sam interrupt other people (more than you would expect for her age), or intrude into their conversations, or into other children’s games?

• A response of "never" was incorrectly treated as the more symptomatic option, inflating the A.3 Developing and maintaining relationships - Adjusting behaviour to suit social contexts score by 11%.

• A response of "often" was incorrectly treated as the less symptomatic option, producing no increase in the subscale score when it should raise it by 11%.

This fix reverses the scoring of those two options to produce corrected subscale percentages.

Note that in the DSM-5 ASD extended assessment, we recommend you always complete every question on the route before generating the report. If you do not, the above errors could be amplified, since questions you did not answer are scored by averaging the answers from those you did.

#115 Count of questions to do is inconsistent

Fixes an inaccurate question count in the "Qs to do" column of the ASD report tables.

In the report, this column describes how many questions were left unanswered on the relevant interview route. Three questions were not correctly counted in this column: q226 (8.8.8), q625 (8.9.7), and q757 (11.6.4). The "Qs to do" count is now accurate.

#122 Remove colour highlighting of ASD table rows

In the Word reports, the tables which report ASD scores in terms of percentages no longer feature yellow highlighting according to built-in thresholds. Instead, users will separately be issued with guidance on how to interpret these scores.

Non-clinical changes

These changes affect the user experience of the 3di software, but not in a way which changes the interview content, scoring or reporting.

#12 Selecting no backend causes unrecoverable error

Fixes an issue where double-clicking the list of backends under Select and open backend caused the application to become unusable until restart.

#75 Use better names for report buttons

To better reflect their purpose, the two buttons which generate Word reports have been re-labelled:

Report (with ASD_extended) → Generate full report

Report ASD_brief → Generate brief report

#77 Update route descriptions

The system route descriptions have been modified slightly for clarity. Following the previous release, there are only seven system routes.

#90 Cannot create report when default Word format has been changed

Fixes an issue where creating a report failed when the Word default file format was set to something other than the standard Word format, .docx.

#103 Remove ResearchPapers from installation folder

The ResearchPapers folder is no longer distributed as part of the 3di installation folder. Find a more up-to-date reading list on our website at https://www.ixdx.org/bibliography.

#104 Remove Tools from installation folder

The Tools folder is no longer distributed as part of the 3di installation folder.

#105 Remove binaries from 3di Resources subfolder

The 3di is now compatible only with 32-bit or 64-bit Office 365 and not previous Office versions. Several program files no longer need to be distributed with the software.

#107 3di installer no longer runs as Administrator

The 3di installer no longer makes any changes at the system level, and no longer prompts to run as Administrator. The default 3di install location remains the user's Desktop folder. For more details, see the installation chapter of the online user guide.

Changes introduced in 250215

Summary

Some non-ASD Word report elements have been removed. The detailed DSM-5 ASD brief scoring table has also been removed.

Written notes and colour highlighting have been added to the Word report for clarity.

Many obscure built-in Routes have been removed.

Detail

1. (#54) Remove table 'DSM-5 scales for SCD, Criteria A and B' from full and brief reports.

2. (#55) Remove the section 'Restricted or Repetitive Behaviours and Interests' which appears in the full report.

3. (#56) There are two tables under 'DSM-5: ASD extended assessment' and two tables under 'DSM-5: ASD brief assessment' which have a column heading including the text 'bigger = poorer'. This text is replaced by '% maximum severity'.

4. (#60) Add colour highlighting to the table under 'Children's communication checklist'.

   1. Rows where the score meets the criterion under column 'Severe deficit if score:' are highlighted in colour.

   2. Scale X receives colour highlighting if the child's score is less than or equal to 130.

   3. Display the <= 130 threshold for Scale X where currently we just have '-'.

5. (#61) Add colour highlighting to the table under 'Attention Deficit Hyperactivity Disorder (ADHD)'. Rows which have a tick in the 'Criterion met' column are highlighted in colour.

6. (#62) The table 'Attention Deficit Hyperactivity Disorder (ADHD)' and the ASD scoring tables currently show unanswered scorable questions in the 'Qs to do' column. This doesn't help the user understand how to fix the problem. Worse, the 'Qs to do' column can optionally be hidden from the Report Preferences dialog, meaning a report can be generated from partial data without any visible indication. If there were any unanswered scorable questions, add a note beneath the table saying 'Insufficient questions were answered to complete this table. Go back and complete the ADHD_brief route' using the appropriate route for each table.

7. (#63) Under 'Tic disorders', the two headings 'Motor tics' and 'Vocal tics' correspond to questions 17.1.2 and 17.1.3 in the interview. Currently the text under these headings says 'None' both when the interviewer recorded that there were no tics present, but also when the interviewer simply didn't fill in the two relevant questions. For each of these two questions:

   1. When the interviewer completes the relevant question and records that no tics were present, the text now says 'None'.

   2. When the interviewer didn't complete the relevant question, the text instead says 'Question [q number] not answered' as appropriate.

8. (#57) Remove most of the routes from the 3di. Only the below are kept: ADHD_brief (DSM-5), ASD_brief (DSM-5 Crit A+B), ASD_brief (ICD-10), ASD_extended (DSM-5 Crit A+B), ASD_extended (ICD-10), CCC.

9. (#58) Under 'DSM-5: ASD brief assessment', remove the pair of tables which give detailed scores in the range 0-6 per behaviour (e.g. Abnormal social approach), leaving just the shorter pair of tables which give scores summarised as percentages per sub-criterion (e.g. A.1 Social-emotional reciprocity).

10. (#59) In all four of the remaining tables under 'DSM-5 ASD', add colour highlighting to all rows with scores strictly over 50%.

Changes introduced in 230122

1. DSM-5 (ICD-11) scoring is promoted, while scoring according to ICD-10 (DSM-4) criteria is available but described as superseded. If you do require ICD-10 scores you can select their inclusion from the Edit report preferences button in the Case Manager. The related interview Routes are available in the Choose route dropdowns.

2. The new DSM-5_brief route and report elements have been added. See .

Changes introduced in 210110

1. SPSS export functionality is completely removed.

2. Support added for 64-bit Office. Compatibility with 32-bit Office retained.

3. No longer includes a bundled Microsoft Access Runtime. This means users are required to have Access preinstalled, or else download the Access Runtime themselves.

4. An early version of the DSM-5_brief subscales (T and U) and routes are present but undocumented. Do not use this feature.

5. The Route Explorer now contains 'jump to question' functionality.

Changes introduced in 190109

Compared to a baseline of version 170107.

1. Bug fix: in certain cases, for example when entering the child's date of birth, dates beyond 2020 had been rejected. Now any date can be entered.

2. Bug fix: in Word reports, a pair of 'Qs to do' cells had been incorrectly swapped in the ASD Outcomes: Developmental normality table. This has been rectified.

3. Bug fix: in Word reports, Chinese characters no longer appear in patient ages in the 3di English edition.

4. Bug fix: previously only the most recently created user route could be selected. Now, any user route can be selected.

5. Bug fix: in Word reports, scores in DSM-5 SCD Criterion B: S1a Functional impairment were incorrectly calculated in certain cases. This has been rectified.