I hear some people say 因为 as “yinwei”, but also some say it sound like “yinvei”. I’m not sure why there is a difference in pronunciation.
/v/ is not an initial found in MSM.
The initial /v/, though, is often found in Northern Mandarin and its branches. In fact, it can still be found, even, in distant branches like Sichuanese.
If you ever saw the 马蜂窝 commercial that aired, ever single night, during the World Cup, you'd certainly notice that 唐僧's "为" is very /v/'d.
You can also refer to the paper The geographic distribution rules and reasons of zero initial “w” pronounced as “v” in Mandarin by Wang Fan. The abstract for the paper goes into some detail in English:
It is universal to pronounce zero initial “w” as “v” in Mandarin. Researches show that 38% of the people with northern dialect do so with 10% in other dialect districts. When analyzing its features and reasons, we find that the dental fricative sound “v” is indeed a variant of the semi-vowel “w”. Its coming into being is affected by the variation of the mouth in different phonemes. It is also related to individual pronunciation habits and the instability of the pronunciation. The universal “v” pronunciation in northern dialect district is caused by the stronger movement of the mouth and the face in pronouncing than that in other dialect districts.
The 为 in 因为 should be pronounced
wei if you're going for "standard" Mandarin. But, if you want to make yourself sound more Northernized you can also opt for the v.
There is no labiodental fricative /v/, including in northerners' pronunciation. What you're hearing is a labiodental approximant /ʋ/, which is sort of halfway between /w/ and /v/. This is a variant pronunciation and phonemically equivalent to /w/.
Because they are phonemically equivalent, both /w/ and /ʋ/ are represented by pinyin "w", rather than "v" or anything else.
The paper cited by the accepted answer looks pretty questionable to me, given that it doesn't make any distinction between labiodental fricative and labiodental approximant. In addition, the translation of 普遍 as "universal" in the abstract is wrong, because pronunciation of /w/ with some dentalization is not universal, but rather a variant pronunciation that (by the paper's own conclusions) is in the minority.