Go to the Experiment pull-down menu and choose Open Spectra. Make sure you select the correct data format at the top of the box – a large number are possible (Azara, Bruker, Felix, NMRPipe, NMRView, ucsf). Change the Experiment Name if you wish (you can always change it again at a later stage, if you want). In general, you should use a new experiment name for each spectrum you read into Analysis. It is in fact possible to read several spectra into one experiment, but this should only be done if the spectra are derived from the same data but processed differently. Change the spectrum name if desired. By default Analysis will read your spectrum into the first possible window. If you have a large number of windows or spectra you may want to change this to All or None instead (though the All option can slow you down if you have a lot of windows). In certain cases you may be using several shift lists – in that case make sure you have the correct one selected (but again this can also be changed at a later stage).
You will now see three pop-up boxes, one after the other. The first gives you information on the matrix size of you spectrum, simply click OK. The second shows you the referencing information. It can be worth checking that this looks reasonable and that the data was referenced correctly when processing it. When reading in Bruker data, for instance, the isotopes may not be defined correctly. Change things if necessary (though again this can be done at a later stage, too).
The third window asks you to select an Experiment Type. Analysis likes to know what type of experiment each spectrum is. It is then able to do intelligent things like automatically know that certain dimensions must belong to the same amino-acid etc. You can narrow down the experiment types on offer by selecting the Category (through-bond, through-space etc.). Then select the name either using the common names for them Type Synonym or the CCPN-defined naming system Full Type. Once you get your head round it, I find the latter easier than the former, as the common names are sometimes different to what I would use. The CCPN naming system essentially puts those atoms which are observed in capitals and those which are used for magnetisation transfer steps in the experiment in lower case. I think the transfers are generally via J-couplings except where stated otherwise (e.g NOESY, TOCSY).