Imagine that we have 10 participants, all of whom would, if our measure were perfectly reliable and valid, score a "4" on our 1 to 7 rating scale. (If you prefer, you may imagine that we have one participant who is really a "4" and is filling out the scale 10 times.)
As you can see from clicking on the 3 buttons below,
  1. a valid measure would also be reliable (a participant who was a 4 would score 4 every time).
  2. a measure that was less reliable would tend to be less valid (if you are getting different scores for the same participant, some of those scores must be wrong).
  3. However, a measure could be reliable but not valid. For example, it would be reliably wrong if it were biased or reliably measuring a different characteristic.