I think this one's a bit of a cliche, but here goes...

In November moved in with a very dear friend: we are very close. Small two-bedroom flat. About two weeks before Xmas, she meets a guy online. Within a week, the other guy has stayed over at our place 3 nights running. This seems very swift and it's a bit surprising for me. Then there was Xmas when everyone was with parents or friends. Then my flatmate stayed over with the new boyfriend but the problem there is that he's a Masters student so lives in a dorm... not ideal for new romance. So then this weekend just gone the new boyfriend stays over again for 3 nights running.

I sense the start of a pattern... it might be that now I live with a couple at weekends... not what I signed up for (and the flat really stretches my budget). Sunday evening I said to my flatmate: "I worry that I'm cramping your style, as you're in the first flush of romance..." i.e. My passive-aggressive way of saying "What is he doing here *again*?" Flatmate looks at me like I'm a cruel alien and says I shouldn't feel that way. She was used to living with her boyfriend and an extra person for many years in the past, so this is her template for a living situation. She looks really hurt and distressed. She is a very loving person and used to having lots of people around. She was also an absolute rock for me during my MSc, so I owe her big-time.

So I would basically like the new boyfriend (who is perfectly pleasant) to be around less. But am I a) sad and jealous cos I miss the closeness with my friend? b) Bitter cos I've been single for AGES? c) in my right mind because the new living situation has been imposed on me and not negotiated?

This friendship means a lot to me, so I don't want to strain it. But it's rather a surprise to be living with a couple at weekends, especially when they only met a month ago. On the other hand, perhaps I should just use the situation as motivation to sort out my own love life? I really am too old for flatshares (I am really really old), and maybe I should just take inspiration and try to get myself married off ASAP...? ,-)

Whaddya think?

Thanks, Killahtron. I did that, and selected the reference category I wanted. But it still insisted on producing output with both reference categories simultaneously! Most odd.

I'm doing a logistic regression and quite a few of my key variables are categorical. I've indicated this and said which I would like the reference category to be. However, in the output, it's including the variable both with the correct reference category (FIRST) and the incorrect one (LAST). Anyone else had this and know how to rectify it?

Thanks, Skig!

I am at least producing models now. Although I berate myself for constantly making foolish errors, I was just thinking that I have developed a bit over the last few years. I used not to be able to touch statistics software without having palpitations and having to follow the instructions in the book word for word, so much so that I couldn't get up to go to the toilet, because it would moving my finger from the text. And I used to become so weird and introverted when doing stats that I would wear the same jumper for days. Now I can do stats and make regular clothing choices, both on the same day.

Good luck all!

Crikey, things are going badly at this end. The recoding of the variables is nightmarish, in a complex survey dataset. The frequencies just don't make sense, aaarrrggghhh!

Am gonna have to abandon that project, and just run models on the inferior, un-recoded variables. :-(

Although I'm *only* a Masters student, the stakes are really high, because I'm on a 1+3. I've had a really tough time academically this year, felt stupid consistently for the first time in my life on this Masters. Compared to my first MSc, where I got a Distinction, this has been absolute hell.

My exam results were mediocre, and the mark I get in this dissertation basically determines whether I get the PhD funding with *my* name already on it, or whether I lose it. So basically how I perform over the next few days until Friday determines the course of my life in a very big way. And I've been under immense time pressure: had to have an extension as the government department that gave me permission to use my original dataset suddenly changed their mind. Bastards.

Anyway, that was a futile vent. Feel free to ignore. But just wanted to let you know things are rubbish at this end too. Misery loves company. :-)

Hello all
Wondered if I could check in today too.

I need to, hopefully by about 8pm

- recode some of my variables
- import into dataset
- remove all respondents who are male
- remove all who didn't answer pregnancy question
- then do multiple imputation
- then recode some other variable
- then run 3 models
- then send models to supervisor

None of these tasks appeared necessary till this morning, so that's a nasty shock.
Sneaks, it sounds really hard for you at the mo. May the force be with you!

Finished the intro! :-)

But now I've remembered I promised to do some proofreading for the journal I'm involved with by tomorrow. So I'll have to do that tonight, and write up my methods stuff tomorrow. :-(

So I'm off to do that for the time being...

Good luck, Batfink! You can do it!

Thanks for welcoming me aboard, Batfink!

It's good to have this thread. I'm now ploughing ahead with my intro, rather than entering my habitual catatonic state.

Hello
Not been on this forum for a long time until this weekend, but was wondering if I could join this thread with you guys. Aiming to work till at least 10 tonight. Finishing my MSc dissertation. It's due in on Friday, and it's been hair-raising for various reasons.

Today I aim to:
- Finish the Intro
- Finish the table describing how I derived my variables
- Finish the text about my variables

Hope it's alright to thread-crash in this way... Who collects the subs round here? :-)

Hi both
Am in a similar position. I have till Friday to do mine, and there's still a fair bit of writing to do. So I'm putting some well-worn notes into the Introduction today. If you're around today we could do some kind of accountability thing. I'll be using tomatoes/pomodoros, so checking in sometime at breaks.

If you're not around - I wish you well! (up)

Hi all
Am having a conundrum with some issues around statistical model selection in the dissertation I'm doing for my 1+3 Masters bit.

I'm getting advice from my friend who's a post-doc at another uni, and he seems confident that doing things his way will help me to do well. But I'm sure that some statistical things he's telling me are things they don't like in my department, where lots of statisticians reside. I'm not getting supervision in my own department, as the deadline for that has now past, so I've no hope of recourse to a real statistician.

The only other person who might advise me is my friend who did well on my course last year and did an excellent dissertation. She is far less experienced than my post-doc friend, and she sometimes goes to him for statistical advice!

My post-doc friend has spent a lot of time helping me, and he gets regularly published, but I'm not sure I can trust his stats. My model selection conundrum feels too complex to be answered with a book.

I realise I'm not making much sense... do others find wildly differing advice offered by different people when it comes to stats? I do. Unfortunately my work is going to be marked by statisticians....

Does anyone know if a non-parametric version of a 2x2 between-participants ANOVA even exists? The things I've been looking up seem to either have only one IV or only one DV. And I have two of both. Do I have a design which can't actually be analysed?

The ANOVA is between-participants.

It seems both of my DVs are positively skewed most of the time (I've checked different groups and levels of the IVs). The DVs are k-value (a measure of impulsivity) and sociosexuality (a measure of how promiscuous someone is/wishes to be). Then the ANOVA IVs are sex (M,F) and exposure to certain stimuli (ladies, landscapes).

But I've collected other IVs like amount of money spent per week on booze, number of hours watching pornography each week, and these are positively skewed. There's one that is negatively skewed and that's religiosity.

I'm inclined to check what a transformation does to the data and then decide whether to go for parametric or non-parametric. Most of the IVs and DVs are crying out for a transformation as they are positively skewed, but as religiosity is negatively skewed it throws a spanner in the works as log-transformation is only for positive skew. You can reverse the scores with negative skew and then log it, but I don't know if I want to mix up this kind of operation in the same analysis with a conventionally log-transformed variable.

Therefore non-parametric tests are looking increasingly appealing...

After rending my garments somewhat yesterday, I have decided that ANY decision must be made today, no matter what.

My sample sizes are great, but there is such large skew on many of my variables I think that ANOVA would be out of the question in this instance. Mostly now I'm just weighing the benefits of a transformation against just going straight for non-parametric tests.

Bootstrapping is beyond my ken - think I will just stick with the (limited) wisdom of Andy Field for now.

Now I'm discovering variables with negative skew. You can obtain reverse scores and log-transform them to make them more normal, but I don't know if you can then have them in the same analysis as variables you have log-transformed in the ordinary way.

I feel I am being lowered down a very deep well.

:-)