aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 11:27:56 GMT
No one is going to shoot you down, aju , when you don't make any assertions about what it definitely shows Thanks for the table! Please can you confirm what the 'weekly volume' refers to? Is it the volume of funds matched in the market (so, including both new lending and RYI transfers)? [Edit: Actually, can this be right? The weekly volume figure is sometimes lower than the RYI figure... hmm...] [Edit Edit: It looks like the average is maybe just an average across all the weeks, updated each week, rather than a moving rolling average? Although I haven't done the sums!] I was being a bit crass on the Rolling bit you are correct I just created a simple function that took the whole average as it went forwards for both. The Weekly Volume is from RS " rate trends" screens where there is a graph showing the relevant rates and volumes etc (Access/rolling is there, Plus and Max are newer and not flagged) The default screen does not have the volumes turned on. I tried adding the two together and just comparing them for each week but gave up. It's interesting that the OP suggests that Rhydian Lewis (RS MD) suggests there is 50% of each but not sure how that fits in with their weekly volume figures. That said I am recording the Weekly volume for all days in week whereas the RYI is only done and reported on Bus Days. You could look here where I am showing them both side by side. I update that daily when I remember.
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 25, 2020 11:36:22 GMT
I was being a bit crass on the Rolling bit you are correct I just created a simple function that took the whole average as it went forwards for both. The Weekly Volume is from RS " rate trends" screens where there is a graph showing the relevant rates and volumes etc (Access/rolling is there, Plus and Max are newer and not flagged) The default screen does not have the volumes turned on. I tried adding the two together and just comparing them for each week but gave up. It's interesting that the OP suggests that Rhydian Lewis (RS MD) suggests there is 50% of each but not sure how that fits in with their weekly volume figures. That said I am record the Weekly volume for all days in week whereas the RYI is only done and reported on Bus Days. Ah yes, this is the screen that I seem to be among the users for whom it's not visible. So, sorry, just to confirm: do we know what this 'volume' is, exactly? And were you suggesting these figures are only for Access? And yes, regardless of what RS have said about the rough proportion of new lending vs RYI releases, I think it's highly unlikely that they would rigidly stick to this every week, or that it's not possible that this proportion would change over time.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 12:40:36 GMT
I was being a bit crass on the Rolling bit you are correct I just created a simple function that took the whole average as it went forwards for both. The Weekly Volume is from RS " rate trends" screens where there is a graph showing the relevant rates and volumes etc (Access/rolling is there, Plus and Max are newer and not flagged) The default screen does not have the volumes turned on. I tried adding the two together and just comparing them for each week but gave up. It's interesting that the OP suggests that Rhydian Lewis (RS MD) suggests there is 50% of each but not sure how that fits in with their weekly volume figures. That said I am record the Weekly volume for all days in week whereas the RYI is only done and reported on Bus Days. Ah yes, this is the screen that I seem to be among the users for whom it's not visible. So, sorry, just to confirm: do we know what this 'volume' is, exactly? And were you suggesting these figures are only for Access? And yes, regardless of what RS have said about the rough proportion of new lending vs RYI releases, I think it's highly unlikely that they would rigidly stick to this every week, or that it's not possible that this proportion would change over time. So the rate trends link (if you are logged in is here. Its a direct link to the page) is very useful for a number of things and since you have visibility to the access product and as an investor you have as much right to the items as any of us I would have thought. If it were me I would be complaining I don't have access to useful information. (I suspect they may be going to pull it at some point I guess when all the 1Y and 5Y products are not available. The figures have always been for all lending its not broken down by product as such on this screen. Or anywhere else to my knowledge. The contents we are discussing is shown in this screen below - click on it for full size.(Hopefully it works for you it goes there for me after I have logged in not before I have.) For those who wish to find this screen (they keep changing things though) then you can do this as follows. 1. Login 2. Select an account 3. Select Market Data in the left menu 4. Select Rate Trends from left menu. In addition to the above picture there are Downloads for daily rates from day one for Access, 1Y and 5Y and is an average market rates for the same 3 products. There is a lot of useful information that RS is limiting if you truly do not have access to this in my view. Another reason for those of that are selling 1Y and 5Y that may lose this if we completely sell out of the older products. This would another clear reason for RS not to want people to know how to understand and select their own rates etc in my view of course.
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 25, 2020 12:55:50 GMT
So the rate trends link (if you are logged in is here. Its a direct link to the page) is very useful for a number of things and since you have visibility to the access product and as an investor you have as much right to the items as any of us I would have thought. If it were me I would be complaining I don't have access to useful information. (I suspect they may be going to pull it at some point I guess when all the 1Y and 5Y products are not available. The contents we are discussing is shown in this screen below - click on it for full size.(Hopefully it works for you it goes there for me after I have logged in not before I have.) Unfortunately, I'm indeed not able to see this page - as I think several other users have stated previously. IIRC, it seemed to be that if you only had access to A/P/M, you couldn't see this? Alas, it's not going to change anything I do - only a matter of interest, so I probably won't bother complaining. Thanks for the screenshot, though! This seems to confirm that it's the total of all funds matched on the market. This is odd, though, given that the RYI release figure is sometimes higher... Unless the weekly volumes figure you're seeing is only for Access + 1 Year + 5 Year (no Plus or Max)? But if A/P/M is one market... Or they don't include the RYI transfers within this figure... Hmmm... As ever with all this, more questions than answers! I should just give up and wait Edit: Thanks for the guide, but the 'Rate Trends' button just isn't actually there, you see
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 13:22:44 GMT
So the rate trends link (if you are logged in is here. Its a direct link to the page) is very useful for a number of things and since you have visibility to the access product and as an investor you have as much right to the items as any of us I would have thought. If it were me I would be complaining I don't have access to useful information. (I suspect they may be going to pull it at some point I guess when all the 1Y and 5Y products are not available. The contents we are discussing is shown in this screen below - click on it for full size.(Hopefully it works for you it goes there for me after I have logged in not before I have.) Unfortunately, I'm indeed not able to see this page - as I think several other users have stated previously. IIRC, it seemed to be that if you only had access to A/P/M, you couldn't see this? Alas, it's not going to change anything I do - only a matter of interest, so I probably won't bother complaining. Thanks for the screenshot, though! This seems to confirm that it's the total of all funds matched on the market. This is odd, though, given that the RYI release figure is sometimes higher... Unless the weekly volumes figure you're seeing is only for Access + 1 Year + 5 Year (no Plus or Max)? But if A/P/M is one market... Or they don't include the RYI transfers within this figure... Hmmm... As ever with all this, more questions than answers! I should just give up and wait Edit: Thanks for the guide, but the 'Rate Trends' button just isn't actually there, you see Not even from the "direct link" I provided, if not I guess they have partitioned off some users from this page, I can't see why though unless they are trying to reduce questions about why newer lenders don't have certain products. The Access/Plus/MAX product are all one queue so I don't think it's partitioned in volumes. You can see from the "Market Data" that they are the same feeds just different positions on the tables. If you don't have the MD screens then you should at least have the invest screens just open two screens with the different products and you will see they are all the same feed just different rate positions. I'm pretty certain that the daily/weekly volumes are one and the same and are for all lending but I've never checked it exactly. (Someone may know better). As for the 50% RYI/Lending. Apart from the fact that my volume figures have two days more data than the RYI info it does seem that if one takes the average of the RYI/(RYI+WkVol) in my tables one gets a rough average of 42% RYI. Now adding in the likely 5/7 for the Volumes then it may not be that far off the mark!. Edit: I tried this in my tables before posting this update - So when I paired back my WkVol using the 5/7 factor and redid the RYI/(RYI+wkVol) and then average the whole list since the start of the wall of RYI in March 16th and guess what the individual averages were up and down but the whole together came out at 50.26% very slightly in favour of the RYI! (Looks like Rhydian promise holds from what I can see)
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 25, 2020 13:31:53 GMT
Not even from the "direct link" I provided, if not I guess they have partitioned off some users from this page, I can;t see why though unless they are trying to reduce questions about why newer lenders don;t have certain products. The Access/Plus/MAX product are all one queue so I don't think its partitioned. You can see from the "market data" that they are the same feeds just different positions on the tables. If you don't have the MD screens then you should have the invest screens just open tow screens with the different products and you will see they are all the same feed. I'm pretty certain that the daily/weekly volumes are one and the same and for all lending volumes but I've never checked it exactly. (Someone may know better) As for the 50% RYI/Lending. Apart from the fact that my volume figures have two days more data than the RYI info it does seem that if one takes the average of the RYI/(RYI+WkVol) in my tables one gets a rough average of 42% RYI. Now adding in the likely 5/7 for the Volume then it may not be that far off the mark. Edit: I tried this in my tables before posting update - So when I paired back my WkVol using the 5/7 factor and redid the RYI/(RYI+wkVol) and then average the whole list since the start of the wall of RYI in March 16th and guess what the individual averages were up and down but the whole together cam out at 50.26% in favoor of the RYI! Haha, no: I cannot access this page. And yes, I'm aware of the market data screens showing it's all one market at different rates; it's purely that your rate trends screen does not have Plus and Max listed (only Access). I agree that it would make most sense for this not to be the case, I'm just trying to work out how the RYI figure could be higher than the volume figure in any week. It simply doesn't make sense that this 'volume' figure is all matches in the market including RYIs if the RYI figure is bigger across a given period (such as one week). Re the 50:50 thing; yes but you're talking about averages. My point was that you can't derive calculations from assuming this split occurs every week. I'm also not really sure taking 5/7 of the volume figure makes sense... We're just talking about the total proportion of market matching that are RYIs. It doesn't really matter what days this occurs on; it matters what the proportion is over a given period. Edit: Your calculation seems to suggest actually that you think that RYIs are not included in the 'volume' total. Is that right? If so, we've been talking at cross-purposes. And is there any particular reason you think this?
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 13:59:58 GMT
Haha, no: I cannot access this page. And yes, I'm aware of the market data screens showing it's all one market at different rates; it's purely that your rate trends screen does not have Plus and Max listed (only Access). I agree that it would make most sense for this not to be the case, I'm just trying to work out how the RYI figure could be higher than the volume figure in any week. It simply doesn't make sense that this 'volume' figure is all matches in the market including RYIs if the RYI figure is bigger across a given period (such as one week). Re the 50:50 thing; yes but you're talking about averages. My point was that you can't derive calculations from assuming this split occurs every week. I'm also not really sure taking 5/7 of the volume figure makes sense... We're just talking about the total proportion of market matching that are RYIs. It doesn't really matter what days this occurs on; it matters what the proportion is over a given period. Edit: Your calculation seems to suggest actually that you think that RYIs are not included in the 'volume' total. Is that right? If so, we've been talking at cross-purposes. And is there any particular reason you think this? I think benaj proved that volume cannot be compared that way by spotting a while ago the discrepancy of RYI as you suggest where RYI is > volume. So since the only other way of collating this is to join them to get a ratio. I'm pretty certain that the Weekly Volume figures are just for lending any way quite apart from that as RYI is a debit not a credit to RS for want of better terms I'm sure there are correct financial one but sadly I'm a software person not a financial one . Therefore since RYI is not lending simply joining them together is a better way to compare them against each other. I agree that using a Weekly Volume Figure and assuming each day is same volume increase to get the 5 day from that figure is flawed so as you say there is an element of error - sadly I cannot go back and redo the figures as the screens for volumes only allows one to do this if monitoring them daily and even then one never knows the correct time to check relative to the RYI and the chart update period. So my theory is flawed at the outset but its odd how close it gets to 50% promise when I ignore individual weeks and deal in the whole period. I'm not saying I'm right but at least I have fiddled with some real figures to an extent and apart from the flaws its odd its so close to Rhydian's commitment. As I said week on week it's not that close but averaging over the whole period its interesting that its so close. All that said and it was fun and if its bollicks then so be it. I'll keep monitoring it that way just because I can.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 14:02:16 GMT
I'm now guilty of muddying an RS thread with a war and peace attempt... I blame the others for the temptation ...
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 25, 2020 14:11:47 GMT
I think benaj proved that volume cannot be compared that way by spotting a while ago the discrepancy of RYI as you suggest where RYI is > volume. So since the only other way of collating this is to join them to get a ratio. I'm pretty certain that the Weekly Volume figures are just for lending any way quite apart from that as RYI is a debit not a credit to RS for want of better terms I'm sure there are correct financial one but sadly I'm a software person not a financial one . Therefore since RYI is not lending simply joining them together is a better way to compare them against each other. I agree that using a Weekly Volume Figure and assuming each day is same volume increase to get the 5 day from that figure is flawed so as you say there is an element of error - sadly I cannot go back and redo the figures as the screens for volumes only allows one to do this if monitoring them daily and even then one never knows the correct time to check relative to the RYI and the chart update period. So my theory is flawed at the outset but its odd how close it gets to 50% promise when I ignore individual weeks and deal in the whole period. I'm not saying I'm right but at least I have fiddled with some real figures to an extent and apart from the flaws its odd its so close to Rhydian's commitment. As I said week on week it's not that close but averaging over the whole period its interesting that its so close. All that said and it was fun and if its bollicks then so be it. I'll keep monitoring it that way just because I can. Hmm... I just wonder if it's somehow more complicated than that. Given that loans within RYI requests will be being 'matched' with investors putting their money 'to buy' in the market (after all, this is the only place the requisite funds come from), it seems odd to me that this wouldn't be recorded as a match in the market. But, indeed, then the figures don't make sense! Given I'm so close to the front of the queue, I wouldn't even care about all this so much normally. But (and I know this sounds like a terrible reason in comparison to others with a load of money at the back of the queue), I have several loans that are just about to go under £10 (when currently I have none), and I'd really love to just have a clean break from the platform rather than having to check back over the course of months/years to withdraw small amounts. But that's just the perfectionist in me.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 14:21:16 GMT
I think benaj proved that volume cannot be compared that way by spotting a while ago the discrepancy of RYI as you suggest where RYI is > volume. So since the only other way of collating this is to join them to get a ratio. I'm pretty certain that the Weekly Volume figures are just for lending any way quite apart from that as RYI is a debit not a credit to RS for want of better terms I'm sure there are correct financial one but sadly I'm a software person not a financial one . Therefore since RYI is not lending simply joining them together is a better way to compare them against each other. I agree that using a Weekly Volume Figure and assuming each day is same volume increase to get the 5 day from that figure is flawed so as you say there is an element of error - sadly I cannot go back and redo the figures as the screens for volumes only allows one to do this if monitoring them daily and even then one never knows the correct time to check relative to the RYI and the chart update period. So my theory is flawed at the outset but its odd how close it gets to 50% promise when I ignore individual weeks and deal in the whole period. I'm not saying I'm right but at least I have fiddled with some real figures to an extent and apart from the flaws its odd its so close to Rhydian's commitment. As I said week on week it's not that close but averaging over the whole period its interesting that its so close. All that said and it was fun and if its bollicks then so be it. I'll keep monitoring it that way just because I can. Hmm... I just wonder if it's somehow more complicated than that. Given that loans within RYI requests will be being 'matched' with investors putting their money 'to buy' in the market (after all, this is the only place the requisite funds come from), it seems odd to me that this wouldn't be recorded as a match in the market. But, indeed, then the figures don't make sense! Given I'm so close to the front of the queue, I wouldn't even care about all this so much normally. But (and I know this sounds like a terrible reason in comparison to others with a load of money at the back of the queue), I have several loans that are just about to go under £10 (when currently I have none), and I'd really love to just have a clean break from the platform rather than having to check back over the course of months/years to withdraw small amounts. But that's just the perfectionist in me. Yeah I struggled with new money coming in but in fact the new money is paying me not RS so technically its not NEW lending as such. That said in the absence of all the useful info and also how RS might call that situation I think I've made quite enough assumptions for one day but it's been fun none the less - Mrs Aju keeps coming round and seeing the blue forum headers and asking me to do other things too - I think I'm happy to be wrong on all accounts and move on before she gets bored and starts online shopping again - that comes out of my budgets not RS's ... Edit: I've just noticed your <£10 loan issue, nothing wrong with your hopes there I have some too but I saw it as an opportunity to keep the 5y for a small fee so punted them in - see another thread if interested!.
|
|
|
Post by diversifier on Jul 25, 2020 15:06:06 GMT
However, the drop in weekly figure below the previous plateau of £3.5m *is* now statistically significant. With two consecutive figures of 2.4, 2.7, we can now be sure that the average is below £3m per week. Going through the (not-very-hard) maths, that shows the level of reinvestment across the platform has now dropped below 50% diversifier , sorry to do this, but as long as you keep making these assertions with such definitive confidence that you're correct, I'm going to say something. Contrast with my posts above illustrating my uncertainty and that I'm piecing bits of evidence together without being sure. Your post is completely statistically illiterate. - Nothing you specify is 'statistically significant'. Statistical significance allows us to reject a null hypothesis at a certain level of confidence based on a statistical test (e.g. t-test, z-test etc.). You seem to just be using it as a meaningless buzzword. - "We can now be sure that the average is below £3m per week" - what average? The average of the last two weeks: yes of course, that's pretty obvious. The average of all weeks so far: absolutely not. Again, what are you even saying here? - How can you know the level of reinvestment so confidently without knowing the levels of loan repayment and levels of new lending for this month? (Just for starters) [Edit: You've now added some calculations. I'm afraid that I fail to follow these also.] t-test? Plot the graph of the 18 weekly figures. It’s reducing during the period before that under investigation, so it needs to be de-trended before one can use the raw data. Standard linear regression: linear trend is -£0.12m per week, Correlation value 0.8. De-trend the data by subtracting that overall gradient. Calculate the population standard deviation of the de-trended weekly figures: £0.45m Let’s exclude the data from weeks 1-7, as we are trying to detect a *second* step-down in RYI amount. The Null Hypothesis is that the true average “recent” weekly (last 3 weeks) is the same as the true average “previous” weekly (weeks 8-15) The linear mean of weeks 8-15 is £3.6m, and the std dev of that estimate is £0.45m/sqrt(8) = £0.16m Recent Weeks 16-18 mean £2.9m, and the std dev of that estimate is £0.26m The means differ by £0.7m, with a combined stddev of £0.3m. *Therefore, we reject the Null Hypothesis at the 2.3-sigma level, which corresponds to p<0.001* Does that help? Please don’t say “I can’t follow your calculations” if you’re going to claim that *I’m* statistically illiterate. It may help you to pause your response, if you know that I have a PhD in quantum physics from Cambridge, and a long senior engineering career in mathematical methods and algorithms.
|
|
benaj
Member of DD Central
Posts: 5,421
Likes: 1,701
|
Post by benaj on Jul 25, 2020 15:25:23 GMT
Like anyone else, My RYI requests are still in the queue since March.
That's okay.
I still manage to withdraw non-RYI money ££££ monthly.
RS is currently my most active platform.
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 25, 2020 15:50:42 GMT
t-test? Plot the graph of the 18 weekly figures. It’s reducing during the period before that under investigation, so it needs to be de-trended before one can use the raw data. Standard linear regression: linear trend is -£0.12m per week, Correlation value 0.8. De-trend the data by subtracting that overall gradient. Calculate the population standard deviation of the de-trended weekly figures: £0.45m Let’s exclude the data from weeks 1-7, as we are trying to detect a *second* step-down in RYI amount. The Null Hypothesis is that the true average “recent” weekly (last 3 weeks) is the same as the true average “previous” weekly (weeks 8-15) The linear mean of weeks 8-15 is £3.6m, and the std dev of that estimate is £0.45m/sqrt(8) = £0.16m Recent Weeks 16-18 mean £2.9m, and the std dev of that estimate is £0.26m The means differ by £0.7m, with a combined stddev of £0.3m. *Therefore, we reject the Null Hypothesis at the 2.3-sigma level, which corresponds to p<0.001* Does that help? Please don’t say “I can’t follow your calculations” if you’re going to claim that *I’m* statistically illiterate. It may help you to pause your response, if you know that I have a PhD in quantum physics from Cambridge, and a long senior engineering career in mathematical methods and algorithms. What you appear to be talking about here is a time-series analysis... And yet you call it a linear regression (which it can't be, because we only have one variable here and regression analyses are for examining the relationship between multiple variables). After this, I'm afraid I really do fail to follow what you're apparently showing here. But maybe, indeed, my own doctoral training in stats isn't up to the level of yours . I only did my undergrad in Tabland, alas... slumming it in London for the postgrad.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 15:57:31 GMT
diversifier , sorry to do this, but as long as you keep making these assertions with such definitive confidence that you're correct, I'm going to say something. Contrast with my posts above illustrating my uncertainty and that I'm piecing bits of evidence together without being sure. Your post is completely statistically illiterate. - Nothing you specify is 'statistically significant'. Statistical significance allows us to reject a null hypothesis at a certain level of confidence based on a statistical test (e.g. t-test, z-test etc.). You seem to just be using it as a meaningless buzzword. - "We can now be sure that the average is below £3m per week" - what average? The average of the last two weeks: yes of course, that's pretty obvious. The average of all weeks so far: absolutely not. Again, what are you even saying here? - How can you know the level of reinvestment so confidently without knowing the levels of loan repayment and levels of new lending for this month? (Just for starters) [Edit: You've now added some calculations. I'm afraid that I fail to follow these also.] t-test? Plot the graph of the 18 weekly figures. It’s reducing during the period before that under investigation, so it needs to be de-trended before one can use the raw data. Standard linear regression: linear trend is -£0.12m per week, Correlation value 0.8. De-trend the data by subtracting that overall gradient. Calculate the population standard deviation of the de-trended weekly figures: £0.45m Let’s exclude the data from weeks 1-7, as we are trying to detect a *second* step-down in RYI amount. The Null Hypothesis is that the true average “recent” weekly (last 3 weeks) is the same as the true average “previous” weekly (weeks 8-15) The linear mean of weeks 8-15 is £3.6m, and the std dev of that estimate is £0.45m/sqrt(8) = £0.16m Recent Weeks 16-18 mean £2.9m, and the std dev of that estimate is £0.26m The means differ by £0.7m, with a combined stddev of £0.3m. *Therefore, we reject the Null Hypothesis at the 2.3-sigma level, which corresponds to p<0.001* Does that help? Please don’t say “I can’t follow your calculations” if you’re going to claim that *I’m* statistically illiterate. It may help you to pause your response, if you know that I have a PhD in quantum physics from Cambridge, and a long senior engineering career in mathematical methods and algorithms. Ah!, Cambridge now those were the days for me and my chums as a youngster at least. Every college has/had a dance/disco/rave/ball best days in my life to be part of the university but not actually be a student!. Bring back the 70's all is forgiven. (I was a towny not a gowny but unless they have changed much I could pin point somewhere for every night of the week and they did little checks that we could not circumvent). Debauchery or what with the the hospital in town, opposite Fitzwilliam museum almost too, instead of on the outskirts as it is now ... Those were the days. Balls (May) were very easy to gate in them days too ... Not sure about your maths thingies though.... sounds too much like hard sums to me! Misspent youth and all that! . Edit: Just remembered the UniChallenge joke when anyone asked us what we read at Cambridge we'd just chime "Newspapers and Magazines"
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 25, 2020 15:59:45 GMT
t-test? Plot the graph of the 18 weekly figures. It’s reducing during the period before that under investigation, so it needs to be de-trended before one can use the raw data. Standard linear regression: linear trend is -£0.12m per week, Correlation value 0.8. De-trend the data by subtracting that overall gradient. Calculate the population standard deviation of the de-trended weekly figures: £0.45m Let’s exclude the data from weeks 1-7, as we are trying to detect a *second* step-down in RYI amount. The Null Hypothesis is that the true average “recent” weekly (last 3 weeks) is the same as the true average “previous” weekly (weeks 8-15) The linear mean of weeks 8-15 is £3.6m, and the std dev of that estimate is £0.45m/sqrt(8) = £0.16m Recent Weeks 16-18 mean £2.9m, and the std dev of that estimate is £0.26m The means differ by £0.7m, with a combined stddev of £0.3m. *Therefore, we reject the Null Hypothesis at the 2.3-sigma level, which corresponds to p<0.001* Does that help? Please don’t say “I can’t follow your calculations” if you’re going to claim that *I’m* statistically illiterate. It may help you to pause your response, if you know that I have a PhD in quantum physics from Cambridge, and a long senior engineering career in mathematical methods and algorithms. What you appear to be talking about here is a time-series analysis... And yet you call it a linear regression (which it can't be, because we only have one variable here and regression analyses are for examining the relationship between multiple variables). After this, I'm afraid I really do fail to follow what you're apparently showing here. But maybe, indeed, my own doctoral training in stats isn't up to the level of yours . I only did my undergrad in Tabland, alas... slumming it in London for the postgrad. On the slopes everyday I bet! unless you mispelt Tabland of course.
|
|