When it comes to research, Google is your friend. :) Do some searching
on window AC units. There are many stories out there. Many are
conflicting...probably because of cheap/defective models and people not
knowing how to size, choose, and use the units. If chosen correctly, a
window AC unit could be your next best friend.

(Even though this is mainly describing window units, much of this also
applies to larger house units.)

Here's a quick recap of the available features on common models:

* EnergyStar: The efficiency of the unit exceeds government standards.
This is always a plus if it doesn't cost that much more.

* Variable Fan Speed Vs. 2 Speed Fans: 2 speed fans are common and good
enough for most people. Variable speed controllers cost more but are
more efficient. If they don't cost too much more than the equivalent 2
speed unit, go for variable.

* Digital Thermostats: These tend to be more accurate than the old dial
type units...at least at keeping the room at a current temperature.
Compared with my mercury thermometer, mine is always a couple degrees
off, though. These are far more convenient at setting a quick
temperature than the guess work of the old dial type units.

* Turn Off Fan: This is usually part of the EnergyStar rating. Once the
unit brings the room down to the desired temperature, it shuts off thus
saving energy. Once every few minutes it will turn the fan on to test
the air temperature. If cooling is needed, it will kick the compressor
on and start running like normal until desired temperature is reached
again.

* Dehumidifier: This is a plus for wet and humid environments. It runs
the unit in slow mode sucking the water out of the air. This seems to be
more sales and marketing as you could just put the unit in slow mode if
needed.

* Soft Starter/Star Booster: When the compressor first starts up, it has a
much higher load. One of these helps boost or split the load relieving
some compressor stress and allowing the compressor to last longer.

* Turn Off Timer: This shuts the unit completely off after a given time
length. I never use it but may be good for running at night then
shutting off in the morning.

* Remote Control: This is for the lazy. In reality, this is good if the
unit is mounted in an area that is hard to reach.

* Quick Slide-In Mount: I wish all manufacturers would use these.
Mounting a 40+ pound air conditioner in a window isn't the easiest thing
to do (especially on the second floor or higher). With these, the light
weight mount gets installed first and the air conditioner easily slides
in afterwards.

Mounting tips: Get some half inch weather stripping and use it all
around the unit. Get some large soft packing foam squares to use as
insulation and use those on each side of the unit to cover the expanding
"wings" to the window edge. Cut a soft foam strip to go under the unit
for draft insulation. Remember that if air keeps blowing in around the
unit, it defeats the efficiency of the unit.

* Noise Levels: Not all AC units sound the same. Some kick in with a
huge thump. Some rattle. Some sound like a freight train. Unfortunately
none of the manufacturers give noise ratings. Search and read reviews
about the model you are interested in. You don't want to install a
thumping freight train model in a bedroom.

* CFM: Cubic Feet per Minute. This is a unit of measure for how much air
flow the blower can do. This is typically listed only for whole house
units (those do 350-450CFM). To get cubic feet of a room, measure the
room in feet then multiply out lenght, width, and ceiling height. An
average room may be 640 cubic feet (8x10x8). If a blower can do 30CFM,
then it will cycle through all the example room's air almost 3 times in
one hour.

* Warranty: Longer full warranty is better. 5 years full coverage seems
to be the longest in the common models. 1 year is normal for the cheaper
units. The longer the warranty the longer the manufacturer believes the
unit will run without any problems. Enough said.

I took another look at the sticker on my window unit. Even though it is
a Hampton Bay model, it is made by LG Electronics (the middle grade
people).

from sticker:
unit=470watts
volts=115v
amps=4.3a
cooling=5200btu
EER=11.0

(In reality the wall voltage can be anywhere from 110v to 120v. Stay
with the sticker math for now.)

Pulling out my power equations, I can dig up the following...

apparent power (for Alternating Current watts) = volts * amps
power = 115v * 4.3a = 494.5VA apparent power

The load is both reactive and resistive, so apparent power isn't really
what it's using. There is a cosine error for the phasing difference
between the volts and amps. Doing a little guess math, mine is about
18degrees off.
470w/494.5VA = 0.95 = cosine(18)
This is probably close for all window units of same size and efficiency.
If I run lower speed, there's probably a little less phasing. Some of
the heavier models could get up to 45 degrees. The smaller the cosine
error, the better.

So your equation starts to come out to the following:
Actual power (watts for AC power) = cosine(phase) * volts * amps
Actual power (watts for AC power) = cosine(18) * 115 * amps

Now, once we get watts (by any means), time to convert to kilowatts:
kilowatts = watts / 1000
kilowatts = 470w / 1000 = 0.47kw

To get kilowatt hours, multiply how many hours per day this unit will be
running.
kwh = kw * hours
kwh = 0.47kw * 4h = 1.88kwh

Look at your electric bill and see how much you are being charged for
electricity. My guess is about $0.13 per kilowatt hour.

dollars = kwh * $0.13/kwh
dollars = 1.88kwh * $0.13 = $0.25

So for running my air conditioner in EnergyStar shut down mode for 8
hours each night (of which about half is where it is turning itself off
= 4 hours), it is costing about 25cents per day. Since I run my AC unit
in low power mode, this will actually be less...the problem is I don't
have any watt or amp numbers to plug into the equations to tell me what
that really is for low power mode.

If I run this unit just like above for a year, cost will be:
year = 365 * $0.25 = $91.25/year
Given that this won't happen during late fall, winter, and early spring,
this number will be quite a bit less than that (probably around $55).

Now that you have the real math behind it all, you can make your own
cost benefit analysis dealing with the efficiencies. Right now I really
wish I had an amp meter to get real world numbers on all this myself.

As far as the sales people at the store go, they are mostly clueless.
They really don't care so they really don't study it much. Generally
speaking, 6months to 2years is about average on cost recovery for a
higher SEER unit. For the next years following until that unit dies, all
money saved is icing on the cake. Also keep in mind that electric bills
are only going to rise over the years.

Since you are going to use your window unit as more of a replacement for
the main house unit for only your room (when the house is empty), you
will be running your window unit a lot more than my booster unit. Your
running costs will be higher per day but you make that up by not running
the far more expensive house AC units during the day cooling a much
larger area. (remember to still run them a few hours after dark to help
prevent humidity and mold issues and to also get the house down to a
more reasonable temperature before sunrise.)

Since you are running the window unit more, the higher the SEER the
better. I think the Frigidaire model we looked at was 10.7 SEER. Your
old unit was a 9.7 SEER. The Frigidaire is 10.3% more efficient than the
old unit. If your main house units are 8 SEER, the Frigidiare is 33%
more efficient. When buying/comparing/swapping/replacing a working unit,
if the new unit is much more expensive and only a few percent more SEER
efficient, the return on investment over time probably isn't worth it.
Keep in mind that the window unit will probably only last 5-10 years and
if the cost of the unit exceeds any savings in the operating cost over
the years, you are taking a financial loss.

Now math gets more complicated here...more so than I can do right now (I
just had my IVIG treatment yesterday and am also feeling lousy). In
short, this involves how fast the AC unit can remove heat per unit time.
The faster it removes heat, the less it runs, thus the less power it
will use overall. This is hard to calculate in human understandable form
so that's why the industry came up with the SEER and BTU/ton numbers.

SEER is a complicated way of measuring BTUs removed per watts used per
hour. To keep it simple, the higher the SEER number, the more efficient
the air conditioner will be. Higher efficiency leads to lower electric
bills and a lower cost of running ownership over time. The older units
are typically 8 SEER. The government mandated higher SEER minimums so
now the base SEER for whole house units is 13. Window units don't seem
to be regulated the same and the max for those seems to be a couple SEER
points less. Unfortunately, higher efficiency units are more expensive
units. The price can be justified if the percentage of operating cost
can be reduced enough. 9SEER is 12.5% more efficient than 8SEER. 10SEER
is 25% more efficient than 8SEER. 10SEER is 11% more efficient than
9SEER. 11SEER is 10% more efficient than 10SEER. Just keep dividing out
the numbers to get the comparisons (Percent_Change=New_Seer/Old_Seer).

Heat removing is measured in BTU's (British Thermal Units). 12,000BTU =
1 ton. So the 6000BTU unit we looked at is a half ton unit. It can
remove 6000BTU's of heat per hour. An average room typically needs
5000-6000BTU's to cool it if there is no other air conditioning present
(this varies a lot on location climate, humidity, sunlight, room volume,
windows, insulation, how many people/computers occupy the room, and so
on).

One thing to watch out for is over capacity and short cycling the AC
unit. If the unit is so big it freezes the air around it, it will run
incredibly short and the other side of the room will never get cold. It
could also form ice and freeze the coils...which is very bad. It won't
cycle the air enough to help remove humidity, either. On the flip side,
an under sized unit will run constantly and won't be able to bring the
temperature down during the hottest parts of the day.

Another thing to watch out for is directable air vents for the cold air
blowing out. One of my units only has side to side. It has problems
sucking back in the cold air and thinking the room doesn't need AC any
more. My other unit only has up and down vents and tends to blow part of
its cold air against the wall. Fully directionable vents are the ideal
choice.

Last thing I can think of to watch out for is the filter tray. One of my
units has a slide in cartridge type that is practically useless. I
cannot do anything else with it. My other one has the open up front
where I can add extra filtering. Extra filtering is more for the benefit
of the AC unit than the people occupying the room. Keeping the fins,
fans, and other innards clean will greatly prolong both life and
efficiency of the unit. Dust acts like extremely fine grain sand paper
and can get in the lubricating grease of the motors slowly wearing them
out faster than normal. Dust can also stick to the moist coils from the
humidity in the air and act like an insulator greatly reducing cooling
capacity of the unit. Over time the wet dust can also turn into a form
of concrete that's near impossible to remove. Usually the black washable
type filter is a good choice to use. It gets the worse of the dust and
doesn't impede air flow that much. Heavier type filters (like HEPA) are
a bad choice for window units because they impede air flow too much.
This causes the fan motor to be overloaded and stressed leading to a
premature failure (which is what we were trying to avoid here in the
first place). A HEPA filter can still greatly benefit the air
conditioner, but put it in a separate air filter unit next to the AC
unit's intake. Also remember the AC unit will not be running constantly
(if sized correctly) so this isn't so good for room filtration which
needs to be always running.

Getting to the point in this example, the Frigidaire is doing the same
amount of work as the old unit, but it is doing it for 10.3% less power.
If the unit runs in low power mode, the SEER rating greatly goes up. In
theory it is possible to be around 20% (or greater) more efficient in
low power mode depending on how the unit is made and operates.