Reading statistics analyses

[Donald Cobb]

Hi all,

 

I'm having trouble getting the information I'm looking for. I'm trying to see how many times in NA28 the expression ἐν (τῷ) ούρανῷ καὶ ἐπὶ (τῆς) γῆς is found, with a possible plural in the first expression (i.e., ἐν [τοῖς] ουρανοῖς). The following works for me:

 

(εν <WITHIN 2 Words> ουρανος) <AND> (επι <WITHIN 2 Words> γη)

 

I tried using the construction module, which I haven't yet gotten my mind around and quickly gave up. I couldn't get it to give me any results at all, or even add the words I was looking for. I wouldn't mind understanding how it works, if anyone cares to show me the steps, but that's not my real question.

 

My question rather concerns the way to read the statistics. It seems that, no matter how I arrange them, they always want to give me the total number of words found. In this case, it comes out to 58. But what I'm actually looking for is the number of times the expression is found. I can manually count (the full expression is used 12 times in the NT according to my count). But how would I need to set the analyses in order for Accordance to tell me that?

 

Thanks for any help you can give!

 

DC 

[Mark Allison]

I think the construct search is really what you want, and it's going to give you the stats you're looking for. Plus, you may want to set the scope to "paragraph" to find some extra occurrences.

 

[Donald Cobb]

Thanks Mark. That helps me with the construct feature. BUT, using the same construct as you give yields 10 results for me, not 13 (or 12, for that matter). For some reason it didn't pick up Mt 18:19 and Col 1:20. I assume that's because of other results in the same chapter (Mt 18:18; Col 1:16), but I don't have scope set to anything other than verse, so I don't see why that should be the case.

 

Trying the same search in BW gives me the same number as my original search, i.e., 12.

 

What is the 13th result that you have found? Here is the list I come up with using the construct search:

 

Μαθθαῖον. 6,10; 16,19; 18,18; 28,18;

Πράξεις. 2,19;

Κορινθίους α ´. 8,5;

Ἐφεσίους. 3,15;

Κολοσσαεῖς. 1,16;

Ἀποκάλυψις. 5,3.13

 

Thanks for your help on this!

 

DC

Edited November 14, 2018 by Donald Cobb

[Λύχνις Δαν]

Hi Mark,

 

  I ran your query and got 16 hits in 37 verses. I tried tweaking search both directions and so on but I don't get your numbers.

 

Hi DC,

 

  The primary difference in your results and Mark's stems from the Verse versus Paragraph scope. Switch to paragraph and you'll either get his numbers or mine. I'm not yet sure why Mark's and mine don't agree.

 

Here are my refs :

 

Μαθθαῖον. 6·9-13; 16·17, 19; 18·18; 28·18; Πράξεις. 2·19; Κορινθίους α ´. 8·5-6; Ἐφεσίους. 3·14-19; Κολοσσαεῖς. 1·11-20; Ἑβραίους. 8·1-2, 4-6; Ἀποκάλυψις. 5·3, 13; 13·6, 8

 

Thx

D

[Donald Cobb]

Hi Mark,

 

  I ran your query and got 16 hits in 37 verses. I tried tweaking search both directions and so on but I don't get your numbers.

 

Hi DC,

 

  The primary difference in your results and Mark's stems from the Verse versus Paragraph scope. Switch to paragraph and you'll either get his numbers or mine. I'm not yet sure why Mark's and mine don't agree.

 

Here are my refs :

 

Μαθθαῖον. 6·9-13; 16·17, 19; 18·18; 28·18; Πράξεις. 2·19; Κορινθίους α ´. 8·5-6; Ἐφεσίους. 3·14-19; Κολοσσαεῖς. 1·11-20; Ἑβραίους. 8·1-2, 4-6; Ἀποκάλυψις. 5·3, 13; 13·6, 8

 

Thx

D

 

Thanks Daniel,

 

Setting the scope to paragraph and using the construct feature, I get the same number of hits as Mark. But several of these are false positives: Mt 6:9 (which only has the first member); Heb 8:1,4 (the two members are there but not in the same paragraph)

 

Furthermore, it doesn't pick up Mt 18:19. In Mt 16:19, it only highlights half the occurrences. In Col 1:20 it highlights ἐπὶ τῆς γῆς, but not ἐν τοῖς οὐρανοῖς. 

 

Looking at your list, Mat 16:17 is a false positive. I would also say that Rev. 13:6,8 are false positives, since they are in separate sentences. 

 

Not sure what needs to be done to improve this. Perhaps putting the two members in the same sentence...

[Λύχνις Δαν]

Yes I noticed the false positives and then I looked at the query again. I think the problem lies in the fact that the query is in two halves, "en ouranw" and "epi ghs". I think there needs to be an additional WITHIN between the two halves to weed some of those out. I didn't try it because Mark's query matched your own and I wasn't clear whether you wanted that additional constraint or not. I am not currently where I can add this and try again but if you add a WITHIN item connected to ouranos and epi and put say 4 or 5 in it that would probably advance things a little.

 

Thx

D

[Mark Allison]

And remember, because I have the search set to paragraphs, Col. 1:20 is a valid hit, since epi ghs is in the same paragraph as Col. 1:16. 

[Mark Allison]

If your looking for results in the same sentence, this is a good option:

 

[Donald Cobb]

Thank you Mark. Are you doing that using the syntax feature? That's the way it looks from here. Unfortunately, I don't have that. But judging from the screen shot, the results from your search don't give the number of occurrences in which the expression is used as such. This was my original question. Or am I missing something? 

 

A couple question about your screen shot: I'm not seeing the other Matt. passages; I assume they are in there but hidden. Is that right? I'm not seeing Acts 2:19 or Col 1:20 either. Are they there?

 

DC

[Mark Allison]

In the search above using the Syntax options, I limited the search to phrases, so it got rid of false hits, but missed some hits that were outside of the phrase boundaries. 

 

Here's another way to perform the search without using Syntax options. I'm ignoring verse boundaries and using a paragraph scope. Because of that, you'll see verses with no hits, because selecting a paragraph scope treats verses as if they aren't there at all. 

 

[Donald Cobb]

Hi again Mark. Thanks for this. But there still seems to be a few problems:

 

1) None of these queries, including the latest one, catches Mt 18:19 (even when I sent the "within" box to 30 words, as opposed to the 15 you had set) or Col 1:20. The reason for this seems to be that the words are reversed in both these verses (e.g., ἐπι τῆς γῆς καὶ ἐν τοῖς ούρανοῖς). I can click on "search in both directions" but that still doesn't catch them.

 

2) There is still a false positive which skews the figures: Mt 6:9 (ἐν τοῖς οὐρανοῖς but not ἐπὶ τῆς γῆς)

 

3) Mt 16:19 has two occurrences; the query excludes the outermost elements (that I underle in the following) and thus takes as one hit the innermost ones (in bold), i.e.: δώσω σοι τὰς κλεῖδας τῆς βασιλείας τῶν οὐρανῶν, καὶ ὃ ἐὰν δήσῃς ἐπὶ τῆς γῆς, ἔσται δεδεμένον ἐν τοῖς οὐρανοῖς, καὶ ὃ ἐὰν λύσῃς ἐπὶ τῆς γῆς, ἔσται λελυμένον ἐν τοῖς οὐρανοῖς.

 

The common denominator between 1) and 3) seems to be that the expressions are reversed. Is there a way around this? As I said, the "search in both directions" doesn't seem to do the trick. Interestingly, my initial search, just using the search line on the menu gives the best results... except for the statistics!

 

Thanks for your patience and you help!

 

DC

Edited November 17, 2018 by Donald Cobb

[Mark Allison]

Hi again Mark. Thanks for this. But there still seems to be a few problems:

 

1) None of these queries, including the latest one, catches Mt 18:19 (even when I sent the "within" box to 30 words, as opposed to the 15 you had set) or Col 1:20. The reason for this seems to be that the words are reversed in both these verses (e.g., ἐπι τῆς γῆς καὶ ἐν τοῖς ούρανοῖς). I can click on "search in both directions" but that still doesn't catch them.

 

 

2) There is still a false positive which skews the figures: Mt 6:9 (ἐν τοῖς οὐρανοῖς but not ἐπὶ τῆς γῆς)

 

3) Mt 16:19 has two occurrences; the query excludes the outermost elements (that I underle in the following) and thus takes as one hit the innermost ones (in bold), i.e.: δώσω σοι τὰς κλεῖδας τῆς βασιλείας τῶν οὐρανῶν, καὶ ὃ ἐὰν δήσῃς ἐπὶ τῆς γῆς, ἔσται δεδεμένον ἐν τοῖς οὐρανοῖς, καὶ ὃ ἐὰν λύσῃς ἐπὶ τῆς γῆς, ἔσται λελυμένον ἐν τοῖς οὐρανοῖς.

 

The common denominator between 1) and 3) seems to be that the expressions are reversed. Is there a way around this? As I said, the "search in both directions" doesn't seem to do the trick. Interestingly, my initial search, just using the search line on the menu gives the best results... except for the statistics!

 

1. Mt. 18:19 isn't found because it's in a different paragraph. If you set the scope to chapter and click "search both direction" it will be found. Col. 1:20 isn't found because it's about 100 words away from the other phrase. And, in my last search, I said there couldn't be an intervening gh, (which there is in the Col. 1:20 instance).

 

2. Mt. 6:9 isn't a false positive because en ouranos is found within 10 words of epi ghs, even though there is an intervening en ouranos.

 

3. The first instance of epi ghs in Matt. 16:19 isn't highlighted, because the search is looking back from the instance of en ouranos it finds in the verse, and because it's searching backwards, sees an intervening ghs, so it doesn't include it in the search. When it searches forward from en ouranos, it finds the instance of epi ghs that is highlighted. That being said, when I remove the constraint that prohibits an intervening ghs, it's still not highlighted. I think that's a bug, and I'm sending it to our developers.

 

 

[Donald Cobb]

1. Mt. 18:19 isn't found because it's in a different paragraph. If you set the scope to chapter and click "search both direction" it will be found. Col. 1:20 isn't found because it's about 100 words away from the other phrase. And, in my last search, I said there couldn't be an intervening gh, (which there is in the Col. 1:20 instance).

 

2. Mt. 6:9 isn't a false positive because en ouranos is found within 10 words of epi ghs, even though there is an intervening en ouranos.

 

3. The first instance of epi ghs in Matt. 16:19 isn't highlighted, because the search is looking back from the instance of en ouranos it finds in the verse, and because it's searching backwards, sees an intervening ghs, so it doesn't include it in the search. When it searches forward from en ouranos, it finds the instance of epi ghs that is highlighted. That being said, when I remove the constraint that prohibits an intervening ghs, it's still not highlighted. I think that's a bug, and I'm sending it to our developers.

 

 

 

Thank you again Mark. 

 

Sorry to be a bother. But what I'm looking for is how many time the pair ἐν [τοῖς] οὐρανοῖς and ἐπὶ [τῆς] γῆς is found together. So I'm looking for a search that won't pick up Mt 6:9, since it doesn't have the second member.

 

I also can't figure out why the search doesn't ever seem to pick up Mt 18:19 and Col 1:20, whether I set the scope to paragraph, verse or whatever. I understand that there may be reasons for explaining why they aren't picked up, but since I'm trying to locate all the occurrences of the expressions together, that doesn't help me. Are you sure it wouldn't have something to do with the expressions being reversed?

 

Again, thanks for your trouble!

 

DC

[Mark Allison]

 

Sorry to be a bother. But what I'm looking for is how many time the pair ἐν [τοῖς] οὐρανοῖς and ἐπὶ [τῆς] γῆς is found together. So I'm looking for a search that won't pick up Mt 6:9, since it doesn't have the second member.

 

 

Not a bother at all. I guess the question is, when exactly do you want to find the pair together? In the same verse? In the same paragraph? 

[Donald Cobb]

Not a bother at all. I guess the question is, when exactly do you want to find the pair together? In the same verse? In the same paragraph? 

 

Within, at most, two sentences, I would say. The main thing is that they are framed as a pair.

[Λύχνις Δαν]

Hi ya,

 

  Getting this perfectly right is tricky. Mt 18:19 is a prime example. When you set the search scope to chapter for Mark's query (search both directions is not required) 18:19 is then found. But it is found as the second part of a result which begins in the previous sentence. But that sentence is concluding the previous paragraph and thought. The two phrases en ouranos and epi ghs in this case are not related semantically. Now whether that is significant for your research I don't know but I just thought I'd mention it in passing. Mt 18:19 cannot be found with this query in its own right because of the reversed direction of the phrases. However, turning on search both directions doesn't help because that reverses the direction of all words not just the two phrases. As a consequence the reversed search is looking for ghs epi ouranos en which doesn't occur here.

 

  If I wanted to find both phrases in either direction I would construct two searches and <OR> them together.

 

Thx

D

Edited November 18, 2018 by דָנִיאֶל

[Mark Allison]

How about this?

 

[Donald Cobb]

Mark,

 

I've been waylaid with a bunch of other activities, but I did want to thank you for this last layout. It seems to be exactly what I was looking for. I'm used to using the GSE (Graphical Search Engine) feature in BW and have been trying to figure out how to do in Accordance what it does in BW. It's taken a little time to get used to the construct feature in Accordance. Your last post helps a lot! Again, thanks.

[Fabian]

I think the construct search is really what you want, and it's going to give you the stats you're looking for. Plus, you may want to set the scope to "paragraph" to find some extra occurrences.

 

 

Mark how do you do this? I can open a second construct search but it is then NOT linked.

 

O.k. I found it out. I have to type it by hand in the search field. 

 

Greetings

 

Fabian

[Fabian]

How about this?

 

 

Mark how do you do this? I can open a second construct search but it is then NOT linked.

 

O.k. I found it out. I have to type it by hand in the search field. 

 

 

It would be fine if we can press again command 1-3 and a new linked construct window opens if a <AND>, <NOT>, <OR>, <XOR> etc. operand is in the search field after the first [LINKED construct] search.

 

Greetings

 

Fabian

[Joel Brown]

Shift-Command-L opens the Link dialog; you rarely have to ever type by hand in Accordance.

[Yohanan]

Hi, I am very sorry to ask such a basic question, but in the new version of Accordance (12.3.1) I cannot find where to get the number of verses or hits in a research tab. It was "naturally" there on the search bar in the precedent versions, I guess there is some pref. somewhere, but where is may question . . .

Yohanan

[Yohanan]

I mean, this is specifically lacking there when selecting a piece of the Bible, but stats analytics appear (only) when proceeding to a word search.

[Joel Brown]

I get the hits and verses at the top of my results: