MattMitchell

Does highlighting the text and using the border icon on the toolbar work?

It's the kind of project that overruns and dies from scope creep, even if it's probably a good idea

@GREED I checked on the rules for this: it's the EO, the Head of Centre (I.e. Head teacher) and anyone else the HoC deems needs to see the results early.

Don't forget that everyone can see any data you've stored in assessment manager aspects though....

If you put all your exams results into exams organiser, it will calculate things like 5 a*-c with e&m, capped points score, etc, all in built-in reports. These are at school and student level, and you can calculate them for any random group from the groups browser too. There's a results in full report too, which may be handy for import to go4schools, depending on the format it likes.

Definitely agree this is something to log with Behaviour - the right-click names thing in lesson monitor makes it easy You can use your own custom codes for lesson, not session, marks - this doesn't affect any statutory returns. Sims maps them using the 'statistical meaning' field.

I'm sure it's there @Cache - it's on the status, not the behaviour I think

In the configuration for the behaviour webpart on SLG, you can choose whether or not to show the comments

You might want a "don't publish on SLG" option too...

That's extremely useful information Phil! I'd always assumed that on the backend of the reporting engine, SQL was generated; you'd expect SQL server to optimise for this kind of thing automatically. Maybe there's actually some kind of intermediate step, where temporary tables are built and then filtered (either by deleting records, or by copying into a new temp table?) As far as indexing goes, if you pull out a dump of the SQL database you can see what those are (or just look at the database itself). Indexing dates can be a bit nasty in SQL server unless you're canny about it, so I can see that the date field might not be indexed. On the other hand, it might just be that the filter on achievement type cut out more records earlier on. Weirdly, I've found on some reports that adding a filter can slow things down! Some "flat" tables are actually views pulling multiple tables together (e.g. student <-> student/achievement relation <-> achievements <-> achievement types ), so being able to miss out certain columns can help with speed too.

Whoops, good point! Had SQL server on the brain at the time, as you can tell

One issue you might run into, depending on what code is in SIMS, is that features in SQL server change with each release, and it may be that SIMS code can't run properly on SQL2003. There are all sorts of little differences between versions, so it's not a safe bet to stick with an unsupported version. Although data security does trump MIS in some ways, not having a guaranteed functional MIS is pretty much a deal breaker these days! Maybe the way round it is to protect the server?

@ICTLady the KS4 calculation sheet is just there to calculate the points score from individual qualifications, but doesn't have to be used for the rest of the exercise. I calculate ours from our exam results output (export results from the "performance tables" report - detailed style), and manually where a student has joined us from elsewhere. You can just enter your KS4 APS right in with the KS5 data, you don't have to use the KS4 sheet if you don't want. Don't forget to include all BTECs (with full weightings), as these *are* used to calculate this KS4 APS figure at present. If you do have the time (and inclination) to calculate value-added on your assessment marksheets, you could use this figure in performance analysis/discover to calculate an average within SIMS. Obviously, this won't be the full value-added figure, as it doesn't take the shrinkage related to cohort sizes into account.

Exactly so! The whole-school published figures are calculated from (sessions attended) / (possible sessions) in total, so a student who is only educated at the subsidiary institution won't have an effect on your overall attendance. I think they'll still be counted in the calculated number of students for working out the percentage of persistent absentees, but they won't count as absent as they won't have missed any sessions!

We use the coefficients from it to track value-added for each data collection. You could actually work out the value-added estimate and difference for a given assessment in SIMS - it's a polynomial calculation, based on KS4 average points per entry (including BTECs, etc): x = KS4 APS estimate (i.e. the value-added 0 point) = c0 + c1 * x + c2 * (x * x) + c3 * (x * x * x) + c4 * (x * x * x * x) (where c0, c1, ... are the coefficients) This figure is subtracted from the assessment value, to arrive at the contribution for that student and grade. The catches that come up are: - the estimated value-added is always capped at the highest possible score for a given qualification type, so that (for example) this won't be higher than A* for an A2 qualification - points are re-based, so that, for example, instead of runnning 0, 150, 180, 210, etc for A2 points, they are converted to be 0, 30, 60, 90, ... This is done so that a failing grade is one grade width below the lowest pass mark - the big catch, is that the value-added for a given subject or qualification type is then "shrunk" by applying the matrix formulae you see in the hidden sheets. Essentially, the fewer students you have for a given course, the more the average value-added is "shrunk" towards zero. A negative figure will always stay negative, and a positive remains positive. I can't think of a straightforwards way of doing this within SIMS, but you could do it in Excel (or SQL server/access/etc) We use the average value-added for interim tracking, as it still conveys the right qualitative message: if student X's value-added for subject Y is lower than you'd like it to be, increasing it will always be a good thing. Also, it's easier to explain the meaning of an average value-added when it agrees with the average of the numbers on a tracking sheet. I've looked at sharing scatter graphs at each data collection, but explaining how to interpret these to teaching staff is quite a challenging task!

I've never tried the AutoReports, but assuming you're producing Excel output via the macros, you might need to check if there's a macro security issue. Just a thought though - if you'd count as a single point of failure, so would sending the scheduled reports only to you! Is it worth looking at setting up a report that more than one person could run (with a simple guide maybe), and have someone run it each day as part of their daily tasks?

I'd be interested to know this too, as we're looking at the same thing here...

Most people tend to have the aspects reflect the subject and type of information (predicted grade, target, attitude grade, etc), and the result set reflect when the data is collected (2013 Summer Term 1, Y10 Autumn 2nd Half, etc). You don't strictly have to do it this way, but it's what most people are used to, and seems to make the most sense!

@LianneKemp just have U set to 0 in your gradeset and it will assign U to anything over and above 0, but less than 80

If you really want to see the structure, the simplest way is to grab a backup of the database and restore it to a new SQL server instance, so you can look at all the various entities and relationships there. To be honest, it probably won't be much use unless you're looking to export/import data in some specific ways, or you're looking to hook into the backend (I'm guessing not recommended without the relevant docs from Capita as these things are always subject to change from version to version). Funnily enough, despite loads of people saying the database wasn't normalised, if anything I'd be tempted to say it looked a little over-normalised, although you're not exactly processing anything like a large dataset in a UK school. The data is often exposed through the reporting interface through views that give a different impression of the underlying tables, or that combine many together (I think I saw a reporting view several versions back that joined about 17 tables together to show information that's presented in the "Student" focus, along with a good few unions...) The other things that could complicate matters for you are that (and I'm going on memory from a few years back now!) some of the functionality that you would imagine is coded into the back end is actually dependent on various values in different tables, especially with some of the group membership stuff; also some schools have a different SQL database in terms of functionality, due to patches that work with specific data, etc. What's more useful, to start with, is to get a good feel for how Assessment Manager works, for how Nova works if you have any involvement with timetabling or curriculum, and for how curriculum memberships work in SIMS. And how all three work together. And how in many schools these, attendance and behaviour, often fall within different areas of responsibility at management level, making a joined-up approach "challenging" at times! Oh, and Exams Organiser is a very strange beast too, and there are exciting opportunities for growth (or whatever the right euphemism is) in working with this and curriculum when it comes to Census time ;-) You might like to look into using Discover, as that can make certain types of analysis a lot simpler, and also other things like Performance Analysis, and simpler tasks like grade distributions from marksheets, using the new Excel-style filter functionality on marksheets, etc. There's a lot to play with, without going anywhere near the backend!

My experience back in more commercial environments was that businesses, sites, and sometimes individual users, often land up using a mixture of date locale settings, so that converting dates (text) to dates (date!) could cause confusion, especially if a date got mangled through UK-text-US misconversion or similar. Rather than deal with people complaining that the dates were coming out wrong, I suspect the chosen solution was to format them as text, so that no-one could complain the wrong data was being output! The "text" format is a pain, though, as it also fails on some versions of Excel when the text is more than 1023/1024 chars long. ironically, formatting this text as "general" avoids the problem. It also leaves Excel to determine that the text date literals are, in fact, dates, and to format them as such. As long as it won't cause any issues at your site, creating a new default template with the format-as-text line removed can make life a *lot* easier, and I've yet to be able to find a test case where it would be a problem.

it's worth getting the results into Exams Organiser if you can, as it will calculate CPS, APS, A*-A, A*-C and A*-G count, etc for you! For Edexcel BTECs, you have to download (or sometimes request to download) the basedata, then you have to submit the registrations by EDI, in order to receive the results by EDI. It's not an instant process, and you may be too late for some (or all) of this year's cohort, I'm not sure off the top of my head. You can add all and any other results manually through the edit PI data option in Exams, although if you have hundreds of results to put in it's going to take a while! There are some seriously helpful improvements to all this scheduled to be included in the Summer release, although whether you want to chance it working or not right on top of results day is another matter. For assessment aspects, I always "normalise" the grade scores to one qualification equivalent (GCSE at KS4 or A2 at KS5) so that, as @vikpaw says, the arithmetic for target and grade comparisons is consistent. "One grade below target" is always a difference of 6 points at KS4 or KS3 and 30 points at KS5 that way. all you have to do then is multiply by the equivalency, so 0.5 for an AS level, 4 for a BTEC L2 diploma (this year at least!) and so-on. Depends on how you like to do your analysis, but I'd strongly recommend setting things up this way, as it makes a lot of stuff simpler.

What would be handy on this one, would be to show marksheets that are linked to the group, for any staff taking the register, rather than just for the actual class teacher. Just a thought...

Thanks for posting this @vikpaw - I missed it on SupportNet and it might save a lot of hassle knowing this now! Not sure how many people import/export results and so-on this way - are there many others out there?

@Stephen_Cleaver with this approach, it doesn't matter that the same subject code is used for multiple actual subjects as much as it might; because you're linking students to classes to marksheets, only those students who appear on a Biology marksheet will have a Biology grade showing. The catch is, of course, that if the same course code is used for multiple sciences, you'd see (for instance) three sets of grade called "science" for a given student, without identifying which is which. I get round this by having a table that remaps the course (and/or subject code and class name) for particular groups, although this is where things vary from school to school. The best way to avoid this, of course, is to use separate codes for different subjects in the timetable, but I know a lot of people timetable all science lessons (for instance) as "Sc" regardless of what subject within Science they are!