You have gathered data on fifty relevant programs and now need to develop a cost estimate. (See below for the 75 observations)

282

355

327

477

343

389

485

312

325

414

391

372

383

463

255

347

461

498

292

431

420

241

441

514

278

305

288

325

383

323

322

267

347

315

432

423

300

354

266

402

378

333

336

288

337

366

284

552

317

292

311

271

578

372

421

383

357

329

330

321

231

463

307

484

371

420

303

299

339

342

340

572

360

276

273

a) Is the apparent distribution plausibly lognormal? How did you come to that conclusion?

b) What are the unit mean and standard deviation? Also, if the distribution is lognormal, what are the transformed mean and standard deviation?

c) Construct a plot (use a lower bound of $100 M and an upper bound of $800 M) of the distribution probability density function (pdf). Indicate the mean value with a plotted vertical line at the appropriate value.

d) What is the coefficient of variation (CV) of this cost estimate?

e) Suppose the program office has budgeted $350M for this program. Plot this budget level on the pdf developed in part c). Given your cost probability distribution, calculate the probability of a cost overrun for the budgeted $350M.

f) The Weapon Systems Acquisition Reform Act of 2009 recommends budgeting to the 80^th percentile of the cost probability distribution. Calculate and plot the 80^th percentile of your cost estimate on the pdf.