Probably the easiest format to import your data is CSV. You can create this in a text editor or a spreadsheet program like Excel or Google Sheets. You can also import an Excel file directly. I haven't used Excel import personally, but some colleagues said it can be a little buggy sometimes. But I suspect it'll work fine for your case which sounds very straight forward.
I'll continue assuming CSV, but the process is essentially the same. This page has some official info.
I assume you already have a coordinate, from a GPS device, for each of your soil sample sites. Organize your excel or csv file so that each row represents one site, and each column is one attribute (eg soil property). Add an additional 2 columns named "X" and "Y" where X = the longitude and Y = the latitude from GPS in decimal degrees (DD not DMS). For example:
site,description,ph,moisture,X,Y
1,"by the tree",7.8,5,-123.3231,45.1234
2,"swamp",6.9,10,-123.3232,45.1235
Open your csv in the project and run the XY Table to Point tool. Here's a description of one way. Note: If your site coordinates are from GPS, they'll almost certainly use the WGS84 coordinate reference system.
That should create a new feature class in the project with your sites, and the soil properties (etc) will be attributes of each point in the new layer.
To perform interpolation of the soil data, probably your best bet is the Kriging tool, but other interpolation methods are available.
