Question re Dates

Gilrucht

Registered User.
Local time
Today, 05:08
Joined
Jun 5, 2005
Messages
132
I realize this sounds like it should be posted in the Report section but I am posting it here because I suspect the answer will involve coding or modifying my forms. I have multiple tables with bound data input forms for my law office db. One of the items recorded is due dates for different things. For example- the statute of limitations for each type of law suit; the date a response to a pleading or motion is due; the date a Court orders filed by, dates of hearings and trials, etc. Each of these duedates is recorded in a different table.
I need to be able to generate a report showing all the duedates for the week. I've struggled with this for a while now. There are simply too many tables to make a query feasable. There are over 25 state civil case tables alone. I need some way to send each duedate to a central storage location for lack of a better term so that I can generate a report. Can anyone offer anysuggestions? Can I send each date to Outlook as it is entered and, if so, how? Or is it possible to create a calendar in Access and record each date in that calendar?
 
Hey Gilrucht,

I guess I could throw a thought your way. What if or can the due dates be put into their own table, tblDueDates and then have each table that needs due dates to relate to it? That way you could have your due dates in one area and then be able to refer to any info in the relating table that you might need. Just a thought. I can't see what your seeing so it may not be doable.

HTH,
Shane
 
This is the result of not having a normalized database. Many such reporting and form problems are, in fact, caused by the same lack of good table design.

Normalization is the process of resolving your data into a set of related tables in such a way that at least the first three rules of normalization are followed.

While there may be work-arounds that you can implement, the most effective solution will be a re-design based on these rules.

For starters, you need ONE table for Civil Cases, not 25.

In this table, you'll also have fields for those details that are common to all suits. Not being well-versed in legal matters, I can only guess at some of the things that would be stored there: Date originally filed, initial jurisdiction, etc. If, as I suspect, you have 25 state civil case tables because there are differences in the requirements in each state, you can retain these tables to enter those details, but instead of the identifying details, you only enter a Foreign Key for the related record in the master case table.

In addition, as has already been suggested, you'll have a table of due dates. THis table will be related back to the master case table, again, with a foreign key. In addition to the date due field, you'll also need a "type of due date" field to indicate, well, which due date is reflected in each record.


HTH

George
 
Last edited:
Actually, George My tables are completly normalized. The state civil cases do not represent 25 different states. They represent different type of civil cases in state court-ie auto accidents, divorces, Medical malpractice, slip and fall, Insurance cases, etc. The db is designed that way to achieve normalization- ie- the data for auto accidents is all related just as the data for divorces is all related just as the data for slip and fall cases is all related. The db has been designed with much help and imput from people here on this forum and is completly normalized.
 
Actually, George My tables are completely normalized
I disagree. I think if you look carefully at the 25 types of civil case tables, you will find many common attributes. And the same goes for all the related tables. You have even more duplicate tables there. In all likelihood, you should have 1 table with a type code and one set of many-side tables related to the single civil case table. With your present design, you will need to make an entire new set of tables to handle a new type of civil case. And what about criminal cases? Do they have their own completely separate set of tables?

An analogy would be a product table for a large retailer like Sears. They sell ladies undergarments and riding lawn mowers and thousands of products in between. Do you think that each of those produces has different attributes? Yes they do and somehow, with proper design, a single product file can accommodate the variations. I think that with some more analysis work, a single file will hold the one-side data for all your cases.
 
I strongly disagree.

Pat's analsysis is right on the mark. Plus, the very fact that you have this type of reporting problem is the strongest kind of evidence that you do not have normalized tables. It simply wouldn't have come up in a normalized table design, because as both of us have suggested, you'd have a table of due dates related to Case, Case Type and the step for which the due date is set.

A Civil Case is a Civil Case is a Civil Case. It consists of a court action between two or more parties. That, in database, terms, is an ENTITY. And each ENTITY belongs in ONE table.

One of the ATTRIBUTES of all Civil Cases is "Civil Case Type".

Auto Accident is one "Civil Case Type".
Slip and Fall is one "Civil Case Type".
Med Mal is one "Civil Case Type", and so on.

You can adequately account for all of your cases with one Civil Case table in which one of the attribute fields is "CaseType". That field is populated with the Foreign Key from the "Civil Case Type" table.

Now, if you still need those 25 additional tables to record specific details for each type of civil case, you can retain them, but you only need the Foreign Key from your Civil Case table to identify them.

George
 
Shane, Thats what I'm thinking but I'm not sure how to do it. I really didn't set forth the scope of the problem and I think that is where George got confused. The db is divided into 4 sub-groups: State-Court Civil Cases. State-Court Criminal Cases, Federal Court Criminal Cases and Federal Court Civil Cases. Within each of these main groups are the specific types of cases. For example in state civil cases there are over 25 different type of civil cases -ie- auto accidents, slip and fall, Workers compensation, Medical Medical Malpractice, Divorce, Wills, Deeds, etc. Each case then has additional tables for info specific to that type of case-ie -there is a childrens table and custody grounds table for child custody; an alimony, property, marriage details, and divorce grounds tables for divorce,etc, accident details, defendants, and auto insurance tables for auto accidents,etc,

Some tables are linked to more than one case . For example, there are witnesss, pleadings, expert witnesses, discovery, etc tables linked to almost all the civil case tables. All told there are probably 150 or so tables in the state civil cases sub-group of the db alone all containing related information so as to achieve normalization. When you factor in the other 3 sub-groups-State Criminal, Federal Civil and Federal Criminal there are close to 500 tables at least half of which have due dates and some have more than one duedate.

To put it in perspective each of the 4 subgroups is treated as a seperate db so I am talking about 4 reports not one but it is still a significant number of dates.
I'd really like to code the after update event to send the duedate to a new table, outlook or an access calendar. That seems like the ideal solution. Can anyone help?
 
Pat, George, Sorry , our posts crossed each other. Here is why I disagree. You are relating civil cases to lawsuits. That is a misunderstanding. Most cases do not end up in litigation. The information recorded for different type of civil cases is completly different from each other. There is no similarity between the information recorded for a client who wants a will written , a deed written, property sold, an estate probated, a contract drafted, their taxes done, a lease written, etc. The only time information will be recorded concerning auto insurance is in an auto accident. The information recorded in a Workman's compensation case is unique to that type of case. The same is true for a divorce or child custody case. The only time I will have information concerning the details of a publication is in a defamation case. The information on a product and manufacture will only be recorded in a products liability case. It is only when the lawsuit is actually filed that there are similarities between the cases . That is handled by having one pleadings table linked to all main case tables; one discovery table linked the same way; the same for witnesses and expert witnesses. But for the most part once you realize most civil cases do not involve law suits I think you can see that that the information is unique to each type of civil case. Even 80 percent of most auto accidents, slip and falls, etc are settled before a lawsuit is ever filed.

Pat, concerning your question re criminal cases, the db is subdividede into 4 subgroups each of which is essentially a seperate db: State civil cases; state criminal cases, Federal Criminal Cases and Federal Civil Cases. Right now I am only working on the state civil cases section of the db. Has this helped?
 
Hey Gilrucht,

Sure, there's a way for that to be done, as far as what your asking for.

Side note: The two individuals that replied to you. Both are what we jokingly call guru's. Both of them spend a huge amount of time helping Access folks on a number of different forums. They both have helped me on more than one occasion over the past few years. At times, their answers may sound like they have an edge to them but bottom line, their helping you not arguing with you. They both very much know what their doing. Mr. George has written a book on this stuff, so I really think that you should consider their advise. My first answer to you was hinting around to you that it sounded like you may have some normalization problems going on. I'm not meaning to dog pile, just stating my two cents worth on not pushing aside George and Hartmans advise.

With that being said. Mr. George did state that otherwise you would be doing work arounds and sure there is a way to accomplish what you would like to do. First thought that hits me is to create your tblDueDate, and in whatever form your wanting the DueDate field value to go to tblDueDate, then you could use an OpenRecord set to take the value from the forms DueDate field and add it to the tblDueDate. You would need foreign keys in the tblDueDate so that it would get assigned to the proper source, so you can refer back to it when needed. Of course, at this point you would become guilty of breaking normalization rules by having redundant data.

I'm sure you will have more to post and we could go from there on what you think.

HTH,
Shane
 
thanks.

I didn't know Mr George but I am very familiar with Pat and I have all the respect in the world for her advice.
 
Pat, Mr. George, In rereading both your posts It just hit me that I think we have a misunderstanding. I think I am already doing what you are suggesting. Its just a difference in labeling.
Mt George writes:
"Auto Accident is one "Civil Case Type".
Slip and Fall is one "Civil Case Type".
Med Mal is one "Civil Case Type", and so on. "

You are right. And I already have that just with a different name. I have

"Auto Accident is one " State Court Civil Case Type".
Slip and Fall is one "State Court Civil Case Type".
Med Mal is one "State Court Civil Case Type", and so on. "

In other words I have a state civil cases table that forms a one to many relationship with all the specific civil casetypes such as autoaccidents, slip and fall, divorces, etc .
I think the issue is what I said in my previous post-the lack of similarity of info between casetypes once you realize few civil cases involve actual court cases
 
Nope, I'm not confused. You have picked out an irrelevant point (Most cases don't go into litigation. Actually I knew that; I didn't think it was important to what you are trying to accomplish here because your question was about due dates related to court actions, wasn't it?) and missed the important point I made.

In fact, you've actualy reinforced my original point, because you've now added a good deal of detail that would have helped the discussion earlier. It turns out that your main cases table is indeed based on the cases your organization handles, which are then related to the detail tables for each type of case--exactly what I said should be the design. You've now added the additional information that you also handle criminal cases and that there are jurisdictional distinctions as well which doesn't change the original principle articulated.

All you need to do is recognize that the reporting requirement you have, which is to be able to aggregate all due dates for all cases, can best be handled in a table for due dates, and that each case, whether it be Criminal or Civil, State or Federal, which has due dates attached to it, can be recorded in that table, along with the type of action for which the due date is needed.


George
 
I have 2 questions. First, Why would I want to put all due dates, criminal and civil, federal and civil, in one table? As currently structured, each of these subgroups-State civil. state criminal. Federal Criminal and Federal Civil are completly independent of each other and there is no inter-relationship whatsoever. For all intents and purposes they are 4 seperate dbs.

Wouldn't I want to follow this structure and have 4 duedates tables-one for each subgroub since we plan on generating 4 seperate reports and keeping each supgroup completly seperate?

Second, It seems I am back where I started when I first posted. I need to determine how to create a table of related data for my due dates table. An example will illustrate the problem. To record the Statute of limitations I need the date of the accident, fall, surgery,contract or publication depending on whether it is auto accident, slip & fall, Med Mal, Breach of Contract or Defamation Action. I also need the clients date of birth because the date is extended for a minor. That is one set of dates.

If a Judge issues a order directing a response by a date certain I need to record the Judge, the date of his order, the due date and the response ordered.. Another due date.

If a pleading is filed I need to record who filed it, what it is and when a response is due. Another due date.

If a deposition is scheduled I need to record who is being deposed, who scheduled it, where it is and when.

If discovery requests are filed I need to record who filed them, what they are and when they are due.

If Medical records or Hospital records are requested I need to record from who and when so they can be tracked.

If a hearing or trial is scheduled I need to record when, where, who the judge is , the time, the purpose, if a hearing, whether or not witnesses have been notified , etc.

I agree I need a duedates table. The problem I am having is that given the dissimilar nature of the information I don't see how they can all be put in one table. I wish they could. Maybe I don't understand normalization but if I put all this unrelated data in one table wouldn't I be violating Normalization principles? Thats why I initially asked if there was a way to send the information to a "central location" from the after update event as it was entered. I understand I would have some redundancy but I don't know that it is unnecessary redundancy.

I'm probably way off base. I just can't conceptualize how to set up one duedates table for all civil cases.
 
Hey Gilrucht,

I'm sure Mr. George and Ms. Hartman will jump in when they get a chance. Until then I'll try to lend a hand. I am very much still learning so I do not understand near like those two, so here goes.

"I have 2 questions. First, Why would I want to put all due dates, criminal and civil, federal and civil, in one table? As currently structured, each of these subgroups-State civil. state criminal. Federal Criminal and Federal Civil are completly independent of each other and there is no inter-relationship whatsoever. For all intents and purposes they are 4 seperate dbs. "

I think what you have said here, states how normalization could be sideways for your db. You are building a "relational database" if you have four "completely independent of each other and there is no inter-relationship whatsoever," then you've stated that the database is not relational cause there is no inter-relationship.

"Wouldn't I want to follow this structure and have 4 duedates tables-one for each subgroub since we plan on generating 4 seperate reports and keeping each supgroup completly seperate?"

The answer here should be no, cause if you do, then your creating redundant data. While the data may be referring to four different instances, the data is the same for all four instances, therefor is redundant. If you keep the "subgroups completely seperate" then your database is no longer relational.

"Second, It seems I am back where I started when I first posted. I need to determine how to create a table of related data for my due dates table. An example will illustrate the problem. To record the Statute of limitations I need the date of the accident, fall, surgery,contract or publication depending on whether it is auto accident, slip & fall, Med Mal, Breach of Contract or Defamation Action. I also need the clients date of birth because the date is extended for a minor. That is one set of dates." :)

Yes, you probably will be back to where you started but not completely back cause you've already indentified quite a few things now, that you didn't know when you started. Why would I know this? Been there, done that. I am also a victim of Mr. George pointing out that my database was not normalized and had to start again. :)

"If a Judge issues a order directing a response by a date certain I need to record the Judge, the date of his order, the due date and the response ordered.. Another due date.

If a pleading is filed I need to record who filed it, what it is and when a response is due. Another due date.

If a deposition is scheduled I need to record who is being deposed, who scheduled it, where it is and when.

If discovery requests are filed I need to record who filed them, what they are and when they are due.

If Medical records or Hospital records are requested I need to record from who and when so they can be tracked.

If a hearing or trial is scheduled I need to record when, where, who the judge is , the time, the purpose, if a hearing, whether or not witnesses have been notified , etc. "

All of the above should really help you to identify what is needed, but when it gets all separated out. It needs to relate to each other.


"I agree I need a duedates table. The problem I am having is that given the dissimilar nature of the information I don't see how they can all be put in one table. I wish they could. Maybe I don't understand normalization but if I put all this unrelated data in one table wouldn't I be violating Normalization principles? Thats why I initially asked if there was a way to send the information to a "central location" from the after update event as it was entered. I understand I would have some redundancy but I don't know that it is unnecessary redundancy. "

You may have already done this but really look at the fields themselves and you should start seeing a similar pattern. While you may have four different instances of what could happen, in the end they do relate to each other. My database I'm building is for our small family owned business, selling medical equipment. We have a ton of different types of equipment to sell, with vendors that sell it and a bunch of ways to get it paid for, but in the end it does boil down to us serving a patient and getting the right equipment, from the correct vendor and paid for (hopefully now days) by the correct funding source. When I sat down and looked at this it seemed like one big cobweb and I was tempted many times to split it all out but working with it long enough got me to see how everything really did relate back to one another.
As far as "maybe not understanding normalization," welcome aboard. It's not an easy subject to wrap your mind around. I've also noticed it's somewhat like a golf swing. There's only one way to hit the ball but there sure are a lot of opinions about how you could hit it better.

Good luck on your project and I hope I helped in some way.
Shane
 
Shane, I don't agree with you at all. There is no relationship between each of the 4 subgroups. Theres no relationship between a state criminal case and a federal civil case for example. We could just as easily have created 4 seperate dbs-one for each of the 4 groups but then we would have to close one to open the other. I don't see where the redundacy comes in???? As I understand what you are saying the information ob due dates wouldn't be repeated. Rather, it would be stored in 4 tables instead of one. Why is that redundant? Even if I were to create only one due dates table we would still generate seperate reports for each of the 4 subgroups. We religously keep all information and paper for each of the 4 groups seperate to the point of storing the physical files for each group seperately. The relationship comes within each of the 4 groups. For the purposes of this question I wasn't even considering the other 3 subgroups to exist. I'm just trying to generate a report of all my duedates for all our state civil cases. I'm not yet working on the other 3 subgroups.
It seemed liked such an innocuous issue when I first posted the question. This is not a new project. I have over 500 tables and have been using this forum for months now. This is the first time anyone has suggested I have a design problem and it is a little disconcerting given how far along I am
 
Gilrucht,

You have every right to not agree and I'm certainly not trying to do anything but lend a hand, not cause you disconcern. I'm not seeing what your seeing so I'm probably way off. It's not fair for me to comment on something that I can't see. Maybe the choice of words on both of our parts has got us missing each other.

As far as posting here and no one ever bring it up could have been due to your questions before didn't shed any light in that direction. I think probably what at least raised the "normalization antenna's" was when you needed to make a report, based on due dates, from various areas, and could not get it to all connect together. That could be a symtom of a normalization problem, so I guess from there the conversation went south for you, in a hurry.

With all of that being said, if I went back to your original question, with out all of the talk in between then I would still come back to my first ideaand that is that you could create a table of Due Dates, have whatever Due Date from where ever, send information to the Due Dates table with whatever information that is needs from the table that sent the Due Date and then build a Report based off of the Due Date table.
 
I am not surprised that no one questioned the structure of your database if you didn't ask any questions that would flag structure as being a problem. Frequently schema design issues don't surface until you try to get "information" out of your tables. You seem to have confused data isolation with normalization. To manage size issues, a spreadsheet solution would be designed very much like this.

At this point, you have invested an awful lot of time and effort, much more than would have been necessary with proper structure. But you have reporting issues and something must be done to the structure or you'll end up in coding hell trying to create reports that cross case types.

The biggest issue with your database is its lack of a Case entity. If you can stand back and look at the structure with a table at the top, some problems will be resolved. The changes I am suggesting are not trivial but you should be able to accomplish them by modifying queries rather than other objects. You will end up with a top level case table which will be the place where caseID is assigned. This table will be the parent to all the existing top level tables and it will become the parent for all child tables that now report to the old top level tables.

An overview of the changes is:

1. Create a Case table. Use an autonumber primary key. Add a case type code and a court code in addition to a civil/criminal type code. Also include a temporary field that will hold the old table's primary key.
2. Move any common fields from the current "top" level tables to this one and remove them from their current place.
3. Change the autonumbers in the top level tables to Long Integer data types. They are going to become the foreign key to the Case table.
4. Create an append query that appends a row for each record of a case table (you'll need 25 for the state civil cases and whatever for the rest). This append will add only the existing primary key to the temp field on the Case table. This process is needed to create the necessary primary keys. If you don't have Cascade Update turned on for all parent-child relationships, do it now.
5. After each new table is added to the case table, it is necessary to propagate its new primary key back to the old children so you now need an update query that joins the new case table to an existing specific case table that updates the old autonumber (which is now defined as Long) to the new key obtained from the new Case table.

Are you confused yet?

6. Once all the foreign keys are fixed, you can merge the data from the duplicate sub tables into a single table. You will ultimately end up with all your transaction dates in a single table. When that is done, you can report on them as a set.

There is an alternative and that is to normalize the data on the fly with union queries. You will end up having to make nested union queries because you have too many tables to union in one query. You can union about eight tables in a single query and then you can union the unions. This isn't pretty and probably won't be fast but it will get all the date data into a single recordset.
 
Pat,
I think I already have done most of that if I understand you correctly. I currently have 4 Main case tables- one for all state civil cases, one for all state criminal cases, one for all Federal Criminal Cases and one for all Federal Civil Cases. Each of the 25(there are actually 38) individual state civil case tables linked to the one main state civil case table. I think maybe I should set forth my structure:

ClientInfo Table (pk = CLIENTid)
ClientFile(PK= ClientID + FileNO. This combined PK becomes the FK for all other tables)(records Fee Information. A combined Key is necessary because aclient can have more than one case)

StateCivil StateCriminal FederalCivil FederalCriminal
---------- ------------ ----------- --------------
Auto Assault Civil Rights White Collar
Divorce Burglary Internet Fraud Stock Fraud
MedMal Theft etc, etc etc., etc.
Contracts etc. etc.
etc, etc

Each case table has additional tables with the details specific to that type of case. There are both business and progamatic reasons for dividing the cases into 4 subgroups like that. First, from a business standpoint, as I told Shane, we religously keep each of the 4 groups seperate in the office . More importantly, Access won't let me link me all the specific casetypes to one main case table. When the total number of cases from all 4 groups , State and Federal, Civil and Criminal are combined they exceed the number that Access will let you link to one table. I tried and got an error message when I reached my 73rd case table. So not only do we want to keep the 4 groups seperate. as far as I can see, we have no choice. We have to keep them seperate and have more than one main case table.
I think what I have to do is to move my duedate fields either to each of the 4 main case tables-StateCivil,StateCriminal,etc or create 4 duedates tables and move them there . Then I can run one report for each group which is what we planned to do all along.
 
Last edited:
Pat,
I reread your Post. I now understand it better. You are suggesting I replace the 4 main tables with one. I think I said this in my previous post but so its clear thats how I started out this db and had to abandon the the idea of one main case table linked to all the individual case types. When I had one main case table at the top I ran up against Access's limitation of 72 links to one table. Thats why I switched to the 4 subgroups- to get around that limitation- that plus this model is more consistent with how our firm operates. So I started out with what you are suggesting but had to abandon it.
 
Gilrucht, I'm not sure that at this point I can help you. Your application is quite complex. You have a great deal of different types of data to store. The analysis required to properly normalize the schema is more than I have time for on an unpaid basis. You also don't seem to be receptive to any massive structural change and that is what I would recommend. See if you can get the union queries to gather all the dates you need to make your reports.
 

Users who are viewing this thread

Back
Top Bottom