EVP recording tips - am I doing this wrong?

I've been experimenting with EVP recording for about three months now and I keep getting frustrated because my results are either completely silent or so full of ambient noise that I can't tell if there's anything meaningful underneath it. I'm using a decent USB digital recorder (Zoom H1, cost me about £70) and I'm recording in fairly quiet indoor spaces - mostly my flat and a mate's old Victorian cottage that's supposed to be haunted.

The problem is that even 'silent' spaces have so much background noise - fridge hum, traffic outside, building settling - that when I go back and listen for EVP, I can basically make anything sound like a voice if I listen long enough. I'm aware this is called pareidolia or whatever, but how do you get around it?

Is there a technique I'm missing? Are people using software to filter out background noise before analysis? Or is this just the nature of EVP work and I need to accept that most sessions will be inconclusive?

You've already identified the main issue - pareidolia is absolutely the elephant in the room with EVP work. The honest answer is that most EVP 'evidence' is filtered through so many layers of confirmation bias that it's nearly impossible to be objective about it. That said, there are some protocols that help: record in pairs, have a second person review the tape independently before telling them what you 'heard', use frequency analysis software to identify voice patterns rather than just listening repeatedly.

Have you tried asking specific questions and leaving gaps for responses? The 'communication' aspect is supposed to be more reliable than random EVP harvesting. Also, dead air isn't helpful - you want to establish context. Ask 'Who are you? What's your name?' then pause, then ask follow-ups. Supposedly this gets clearer responses than just leaving a tape running.

Audacity is your friend here. Free software, proper audio analysis tools. You can isolate frequency ranges, apply noise reduction filters, zoom in on specific sections. But honestly - and I say this as someone who's spent actual money on this hobby - the EVP field is so plagued with confirmation bias that even with good technique, you're fighting an uphill battle against your own pattern-recognition instincts.

The Victorian cottage in Staffordshire? If it's the one I'm thinking of, someone's done some investigation there before. Worth checking if there's any documentation online about what's been recorded there previously. Sometimes knowing the history helps you ask better questions or at least understand what 'activity' supposedly happens at specific times.

Is there a technique I'm missing?

The Zoom H1 is decent kit. The issue isn't your equipment, it's that EVP as a discipline fundamentally lacks good methodology. Try this: record in the location, then take the recordings home and listen without any expectations or preconceptions. Don't tell yourself 'I heard something', just listen genuinely. You might surprise yourself with how quiet most sessions actually are.