I have done backtesting for a set of symbols and get two different values depending on which method I use.

Method 1:
1. Doing an optimization with two different optimization parameters.
2. After finishing the optimization, in the Results section click "View Average Results" section.
3. Results for a particular set of values for the optimzation parameters a Net Profit of $256.

Method 2:
1. In the strategy editor window, set the same values for the optimzation parameters as from step 3 above.
2. On the dataset tab click "Backtest on all ..." button.
3. In the results report the net profit displays as $12,000

For both methods I have the same values set for scale, data range and Position Size.

Does anyone know why there are two very different results from the two methods?

You're running in Portfolio Simulation mode and not setting Position priorities.

See User Guide > Strategy Window > Backtesting Strategies > How Trades Are Chosen.

Hint: try starting from the FAQ and Errors section when you have a question or an issue. The same answer exists on our FAQ for years:

Every time I run a Strategy I get a different result. What am I missing?

I have seen the FAQ about Portfolio Simulation and how trades are chosen.

For both methods though, in the upper left corner of the screen, in Position Size, I have Fixed Dollar, under Raw Profit Mode picked with the default value of 5000.

It is for that reason that I am confused about why I am seeing two different results for Raw Profit Mode, not that I'm using Portfolio Simulation and have not read the FAQ carefully.

If one of the methods I mention above automatically uses Portfolio Simulation even though I have Raw Profit mode selected that would be useful to know (and information I have not seen in the User Guide or the FAQ).

Good point: RP mode makes difference.

However, you're comparing apples to oranges. "View Average Results" is averaging the result of separate optimizations for all symbols of the DataSet whereas backtesting on all symbols (even in RP mode) creates a portfolio backtest (although not true to real-world rules because position sizing is still RP). Hence the difference.

In other words, the $256 Net Profit from an optimization on e.g. Dow 30 is based on averaging the 30 individual net profit figures. The $12,000 figure comes from a single portfolio backtest in RP mode.

Hope that makes sense.